Posters
281
Proceedings of the Sudden Oak Death Fourth Science Symposium
Status of California’s Phytophthora
ramorum Stream Survey1
Kamyar Aram2 and David M. Rizzo2
Abstract
Since 2004, we have annually monitored Phytophthora ramorum in streams, creeks, and
rivers throughout infested and at-risk areas of California by baiting with rhododendron leaves.
This survey helps delineate the pathogen’s range and can quickly detect its spread. Our sites
are dispersed from Del Norte to San Luis Obispo counties on the coast, and in Butte, El
Dorado, Nevada, Placer, and Yuba Counties of the Sierra Nevada foothills. Monitoring runs
from February through June, the season of mild and moist conditions and thus highest
pathogen activity.
We deploy rhododendron leaf baits of mature cuticle encased in fiberglass mesh bags into
streams for 7 to 21 days, starting with the longest interval for the early cold season, and
decreasing bait exposure time as the season warms and stream temperatures rise and cause
leaves to degrade more quickly. Baits are deployed once per month from February through
June. At the end of each monitoring period, bait leaves are retrieved, surface sterilized, and
cultured on Phytophthora-selective media. P. ramorum and other Phytophthora species are
identified visually using a microscope, and in certain cases, confirmed by DNA analysis of
bait tissue or cultures.
New detections of P. ramorum in 2008 occurred primarily adjacent to previously known
infestation centers such as Redway, Humboldt County, the Navarro River watershed,
Mendocino County, and Big Sur in southern Monterey County. A new detection in the Little
River at Van Damme State Park marked the northern-most report of the pathogen in
Mendocino County at the time, and was the most distant case from a previously confirmed
occurrence. No terrestrial inoculum source has yet been identified for this site or infested
creeks in the city of McKinleyville, Humboldt County. To date in 2009, we have only
detected the spread of P. ramorum westward from previously confirmed sites near Redway in
Humboldt County while detection of the pathogen continued for most sites confirmed in
previous monitoring years.
In 2008, the second relatively dry season in a row, the pathogen was cultured predominately
from February and March samples. The most well-established P. ramorum-infested sites (for
example, the Navarro River watershed) had the most consistent detection, while a few with
previous detection were negative. In 2009, a season with more regular rainfall, the pathogen
was detected at some sites into May, but nonetheless the total number of detections again
declined sharply after March. This seasonal variation suggests that occurrence of P. ramorum
in streams coincides with periods of pathogen activity, and that detection in waterways during
dry periods may be diminished. Nonetheless, the pathogen’s detection in both new and
previously confirmed waterways indicate that this is a sensitive method for monitoring the
spread of P. ramorum.
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
University of California at Davis, Department of Plant Pathology, One Shields Avenue, Davis, CA.
283
GENERAL TECHNICAL REPORT PSW-GTR-229
A portion of this work is part of the national P. ramorum stream monitoring program
supported by the U.S. D epartment of Agriculture, Forest Service. This project involves many
collaborators whose assistance, permission, and guidance make this work possible.
284
Proceedings of the Sudden Oak Death Fourth Science Symposium
Sampling Cankers and Lesions on Eastern
Forest Plant Species for Phytophthora spp.
and Associated Organisms1
Yilmaz Balci,2 William L. MacDonald,3 and Kurt Gottschalk4
Introduction
Our studies, begun in fall 2003 in eastern and central U.S. oak forests, have
demonstrated that native, exotic, and new Phytophthora species are common to forest
soils (Balci and others 2007, 2008). In sites where we have isolated Phytophthora
from rhizosphere soil samples, collected around the base of oak trees, usually no
aboveground cankers or lesions are found. However, some of the species found in our
survey sites such as P. cinnamomi, P. cambivora, and P. citricola can cause bleeding
cankers as well as foliar infections on species commonly found in eastern U.S. forest
ecosystems (Erwin and Ribeiro 1996). Whether in Phytophthora-infested sites any of
the cankers or lesions can be associated with Phytophthora remains uninvestigated.
Thus we surveyed trees and understory plant species for any symptoms that resemble
Phytophthora infection and performed isolations to detect and evaluate the incidence
of this group of organisms. Other commonly associated organisms were also
identified.
Material and Methods
Ten sites were chosen from the 2004 multi-state Phytophthora survey that proved
positive for Phytophthora (Balci and others 2007). Sites with highest isolation
frequencies and species diversity were selected. Each survey site consisted of a
circular 50 m radius. At each study site, a survey of trees and shrubs was undertaken
to examine stems for any bleeding or canker development. Such trees were examined
for fresh expanding necroses by removing the outer bark layer. Several bark and
attached wood samples from the outer edge of an expanding lesion were taken from
each tree and placed in plastic bags. Before the isolation process, samples were
washed several times to remove any oxidative staining and polyphenols. Small pieces
of the necrotic inner bark tissues then were plated directly on a Phytophthora
selective media (PARPNH) using half of the wood chips collected (Erwin and
Ribeiro 1996). No surface disinfections were performed prior to plating of samples
on selective media. A portion of the other half of the chips was plated on malt
extract agar (MEA) containing 100 mg L-1 streptomycin sulfate to evaluate the
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
University of Maryland, Department of Plan Sciences and LA, 2114 Plant Science Build. College Park,
MD 20742.
3
West Virginia University, Department of Plant and Soil Sciences, 1190 South Ag. Build. Morgantown,
WV 26506.
4
USDA-Forest Service, Northern Research Station, Morgantown, WV 26505.
285
GENERAL TECHNICAL REPORT PSW-GTR-229
most frequent organisms associated with canker formation. Surface sterilization prior
to plating was as follows: wetting 1 minute in ethanol (96 percent), surface sterilizing
for 5 minutes in sodium hypochlorite (4 percent), dipping for 30 seconds in ethanol
(96 percent), rinsing in sterile deionized water, and drying on sterile filter paper.
At each site, leaves of understory plants also were surveyed for lesions that may
represent Phytophthora infections. Suspect leaf samples were wrapped in paper
towels in the forest and transferred in plastic bags to the lab. Sections of necrotic
leaves were plated onto Phytophthora selective media (V8-PARPNH) as well as on
malt extract agar after surface sterilization as described earlier.
Although isolations were primarily conducted from necrotic bark tissues, soil
samples as well as roots were collected from around sample trees in order to evaluate
the resident soils for the occurrence of Phytophthora species. In addition to the five
oak trees sampled previously, five other tree species were sampled. Soil samples
were tested for Phytophthora with an oak leaflet baiting technique as described in
Balci and others 2007. Roots collected from each of the soil monoliths were washed
with pressurized water, and any necrotic tissue processed similarly as the canker
samples found on stems.
Results and Discussion
Some of the significant findings were as follows:
 Soil samples collected in 2006 yielded similar isolation results as in 2004 for P.
cinnamomi. However, this was not true for P. europaea and P. quercetorum,
suggesting fluctuation of population of different species over the years (table 1).
 In sites infested by P. cinnamomi, other tree species were also found to be infected
by this species. In various West Virginia sites, beside the oak species, soils
collected from the following plant species also resulted in positive isolation: Acer
rubrum, Aesculus octandra, Carya sp., Fagus grandifolia, Liriodendron tulipifera,
and Nyssa sylvatica.
 On coarse roots extracted from oak trees, multiple small lesions with a distinct
margin were found. These lesions resulted only in few incidences with isolation of
P. cinnamomi (table 1). The isolation success was very low and limited to one tree
at a sampling site. In one site at Ohio, necrotic fine roots also resulted with P.
cinnamomi isolation.
 Direct plating of root lesions on malt extract agar resulted with isolation of
multiple fungal isolates. Among the isolates, two commonly isolated species,
Cryptosporiopsis and Cylindrocarpon, are known to be pathogenic. Other
potential pathogens include Colletotrichum cf. trichellum and Cylindrocladium sp.
 Lesions found on Rhododendron maximum and Kalmia latifolia were similar to P.
ramorum infections found in nurseries. None of the necrotic leaves found on
Hamamelis virginiana or Magnolia sp. resembled Phytophthora infection. In no
instances was Phytophthora isolated from any of the foliar samples. Among the
fungi isolated from leaves, Pseudocercospora sp. and Pestalotiopsis sp. were the
most commonly isolated fungi on foliage of kalmia and rhododendron,
respectively. These two species are known to be pathogenic on foliage of a variety
of other plants.
 Most of the stem cankers with tarry exudates or bleedings could be classified as
caused by insect damage, bacterial bleeding, frost, or wood-decay organisms.
286
Proceedings of the Sudden Oak Death Fourth Science Symposium
However, in some incidences they closely resembled Phytophthora lesions. Despite
the similarity of those lesions, no Phytophthora was isolated from any cankers in the
study sites.
Table 1—Phytophthora spp. isolated from different plant parts and soil
samples in 2006. All of the study sites were infested with a Phytophthora
species, which were determined based on isolations from rhizosphere soil
samples collected from oaks in 2004 (Balci and others, 2007)
Acknowledgments
This research was made possible by funding from the U.S. Department of
Agriculture, Forest Service, Pacific Southwest Research Station (05-JV-11272738089). Special thanks are given to Selin Balci for her help in laboratory and field
routines. We would also like to thank Jordan Eggers, Jill Rose, and Drs. R. P. Long
and T. J. Hall for their assistance during extensive field work.
Literature Cited
Balci, Y.; Balci, S.; Eggers, J.; MacDonald, W.L.; Juzwik, J.; Long, R.P. and Gottschalk,
K.W. 2007. Phytophthora spp. associated with forest soils in eastern and north-central U.S.
oak ecosystems. Plant Disease. 91: 705–710.
Balci, Y.; Balci, S.; Blair, J.E.; Park, S-Y.; Kang, S. and MacDonald, W.L. 2008.
Phytophthora quercetorum, a novel species isolated from eastern and north-central USA oak
forest soils. Mycological Research. 112: 906–916.
Erwin, C.D. and Ribeiro, O.K. 1996. Phytophthora diseases world-wide. St Paul, MN:
American Phytopathological Society Press. 562 p.
287
GENERAL TECHNICAL REPORT PSW-GTR-229
An Evaluation of Phytophthora ramorum
Preventative Treatments for Nursery and
Forest Understory Plants in British
Columbia, Canada1
Elisa Becker,2 Grace Sumampong,2 Simon F. Shamoun,2 Marianne
Elliott,3 Aniko Varga,4 Saad Masri,4 Delano James,4 Karen Bailey,5
and Susan Boyetchko5
Abstract
Foliar treatments were applied to potted Rhododendron sp. cv ‘Cunningham’s white’ plants
over a 3-week period and evaluated for their ability to protect the leaves from Phytophthora
ramorum infection. Treatments included several chemical and biological control products and
systemic acquired resistance (SAR) elicitors. Excised leaves from treated plants were
challenged by application of P. ramorum sporangia. Our preliminary results indicated that the
most effective treatments were the two products registered for prevention of P. ramorum in
Canada, Aliette® and Subdue Maxx®. Biocontrol treatments Rhapsody® and Actinovate® and
SAR elicitors 3-aminobutyric acid (BABA) and Actigard® conferred a lesser degree of
protection to the leaves and lowered the incidence of P. ramorum infection. Treatment with
Sonata® was not significantly different than that with water. The chemical BABA did not
inhibit P. ramorum growth in vitro. Plant trials are ongoing.
Introduction
On the west coast of British Columbia, Canada, several nurseries have reported
Phytophthora ramorum-infected plants, but we have no indication whether the
pathogen has spread beyond these points of entry. Plants in our forests and wildlands,
as well as in our nursery industry and gardens, however, are vulnerable to this
pathogen. As part of a collaborative research project among the Canadian Forest
Service, Pacific Forestry Centre, the Canadian Food Inspection Agency and
Agriculture and Agri-Food Canada (AAFC), we are evaluating treatments that may
protect susceptible plants from infection, to be included in an integrated management
approach.
Several commercially available biocontrol products, Rhapsody® and Sonata®
(containing Bacillus spp.), and Actinovate® (Streptomyces sp.), were previously
tested in vitro by using dual culture and treatments to detached leaves (Elliott and
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
Canadian Forest Service, Pacific Forestry Centre, Victoria, BC, Canada V8Z 1M5.
3
Washington State University, Puyallup Research and Extension Center, Puyallup, WA 98371.
4
Sidney Laboratory, Canadian Food Inspection Agency, Sidney, BC, Canada V8L 1H3.
5
Agriculture and Agri-Food Canada, Saskatoon Research Centre, Saskatoon, SK, Canada S7N 0X2.
Corresponding author: elisa.becker@nrcan-rncan.gc.ca.
288
Proceedings of the Sudden Oak Death Fourth Science Symposium
others 2008, in press). In the present study, we applied these treatments to whole
plants and evaluated the response to P. ramorum infection on excised leaves of the
treated plants. This plant trial also included treatments with chemical agents that may
induce systemic acquired resistance (SAR) responses: Actigard® (benzothiodiazole),
and 3-aminobutyric acid, (BABA) (table 1). We also included the chemical
fungicides Subdue Maxx® (metalaxyl-m) and Aliette® (fosetyl-A), which are
registered for use on P. ramorum host plants in Canada.
Methods
Plant Trials
Host plants were 1.5-year-old rooted cuttings of Rhododendron catawbiense cv.
'Cunningham’s white' in gallon pots. Plant health was assessed before and after the
trial. Eight foliar spray treatments were applied, as per manufacturers’ label
instructions or as noted for BABA (table 1), to five plants each. All treatments were
applied once-a-week for 3 weeks, except for Aliette® and Subdue Maxx®, which were
applied on a 2-week interval. Seven days after the last application of all treatments,
12 leaves per plant were excised and brought to the lab for challenge with P.
ramorum. Sporangia were produced by growing P. ramorum isolate PFC 5073
(RHCC23), lineage NA2, in V8 broth for 4 days, then cultures were washed and
media was replaced by water for 2 more days to induce sporangia formation. Cultures
were poured through cheesecloth and the resulting purified sporangia solution was
diluted to 10 000/ml. Half of the excised leaves from each treated plant (six) were
sprayed with P. ramorum sporangia. All leaves were incubated for 14 days in sealed
plastic boxes containing moist vermiculite, then disease response was assessed by
counting the number of leaves with lesions as a percent of total leaves challenged
with P. ramorum per treatment.
Table 1—Active ingredients and rates of foliar treatments to plants
Name, source
Active ingredient
Trial rate
Actinovate® SP,
0.9g/litre
Streptomyces lydicus
Natural Industries, Inc.
* added Silwet L-77
Rhapsody® ASO,
20ml/litre
Bacillus subtilis
Agraquest
Sonata® ASO,
10ml/litre
Bacillus pumilus
Agraquest
Actigard®,
benzothiadiazole
0.09g/litre
Syngenta
BABA,
3-aminobutyric acid
1g/litre
Aldrich
Aliette® WDG,
fosetyl-A
5g/litre
Bayer
SubdueMaxx®,
metalaxyl-m
156μl/litre
Syngenta
Water
PPM
901ppm +
156ppm Silwet
20,000
10,000
90.1
1000
5006
156
289
GENERAL TECHNICAL REPORT PSW-GTR-229
Dose Response In Vitro
To evaluate the response of P. ramorum to BABA, 200μl cultures were initiated from
1000 zoospores per well in 96-well plates. Final concentrations of BABA ranged from
0.1 ppm to 1000 ppm. Plates were incubated at 20 oC. To estimate the growth of P.
ramorum cultures, the optical density (OD) (650 nm) was measured every 24 hours for
3 days.
Results and Discussion
Our preliminary results of one plant trial indicated that the most effective foliar
treatments for the protection of healthy rhododendrons from infection by P. ramorum
were the two chemical fungicides registered for this use in Canada: Subdue Maxx® and
Aliette® (fig. 1). Treatment of plants with the SAR elicitors BABA and Actigard® and
biocontrol treatments Rhapsody® and Actinovate® resulted in some degree of protection
to the leaves against P. ramorum and reduced the number of foliar lesions. Treatment
effects of the biocontrol Sonata®, were not significantly different than treatment of
plants with water (fig. 1). No treatments had any apparent phytotoxic effects. More
assessments were made of this experiment than are presented here, including size of
lesions and effects of leaf age. These data will be presented along with the results of the
repeat of this experiment.
100
% Infection
80
B
60
40
B
AB
AB
AB
AB
20
A
A
Water
Subdue
Maxx
Sonata
Rhapsody
BABA
Aliette
Actinovate
Actigard
0
Figure 1—Disease response on excised leaves of treated plants. The number of leaves with lesions is
given as a percent of total, per treatment. A one-way ANOVA comparing treatment means was
performed. Treatments with the same letter were not significantly different according to the Tukey test at
P = 0.05.
The dose response of P. ramorum to BABA was assessed in vitro over a wide range of
concentrations. There was no significant difference in growth after 3 days among the
cultures in different concentrations (0.01 to 1000 ppm) of BABA, showing equivalent
results to growth in 0 ppm (P = 0.165) (data not shown). The chemical BABA did not
290
Proceedings of the Sudden Oak Death Fourth Science Symposium
inhibit the growth of P. ramorum in liquid culture in concentrations up to 1000 ppm.
This was the concentration used in foliar spray treatments (table 1).
More plant trials are underway on rhododendrons and other nursery and forest plants
known to be hosts to P. ramorum. Root treatments are also being tested. One objective
of these trials is to test whether resistance responses in treated whole plants can be
detected in excised leaves challenged with P. ramorum. Infection of intact leaves on
treated plants by P. ramorum will be performed in AAFC containment facilities and
will be compared with the infection of excised leaves. Further in vitro screening is
ongoing to identify other candidate treatments with suppressive effects against P.
ramorum growth, including other microbials, plant extracts, and surfactants, as well as
disinfectants for greenhouse use.
Literature Cited
Elliott, M.; Shamoun, S.F.; Sumampong, G.; James, D.; Briere, S.C.; Masri, S. and Varga,
A. 2008. Identification of control agents and factors affecting pathogenicity of Phytophthora
ramorum. In: Frankel, S.J.; Kliejunas, J.T.; Palmieri, K.M., tech. coords. 2008. Proceedings of
the sudden oak death third science symposium. Gen. Tech. Rep. PSW-GTR-214. Albany, CA:
U.S. Department of Agriculture, Forest Service, Pacific Southwest Research Station: 347–350.
Elliott, M.; Shamoun, S.F.; Sumampong, G.; James, D.; Masri, S. and Varga, A. [In press].
Evaluation of several commercial biocontrol products on EU and NA populations of
Phytophthora ramorum. Biocontrol Science and Technology.
291
GENERAL TECHNICAL REPORT PSW-GTR-229
Phytophthora ramorum Recovery from
12 Big Sur Watersheds Following Wildfires
in 20081
Maia M. Beh,2 Kerri Frangioso,2 Akiko Oguchi,2 Kamyar Aram,2
and David M. Rizzo2
Abstract
Many regions of California affected by sudden oak death (SOD) are prone to fire. Fire is an
influential natural process in forest ecology, yet there have been very few studies on the
effects of wildfire on forest pathogens, especially invasive pathogens. In the Big Sur area,
where forests are among the most impacted by SOD, large areas were affected by fires during
the summer of 2008, with over 240,000 acres burned in the lightning-sparked Basin-Indians
Complex Fire and another 16,000 in the Chalk Fire. Field plots established in Big Sur in 2005
to examine the feedbacks between Phytophthora ramorum and the physical environment have
provided information on pre-fire disease incidence, tree mortality, and various other
biological and physical forest characteristics. Following the fires, burn severity, an evaluation
of fire intensity, was also assessed using remotely-sensed as well as on-site plot
measurements. Our pre- and post-fire data allow for the unique opportunity to study the
interactions between SOD and wildfire. Some of the questions our lab is investigating
include: Did SOD make the fires burn more intensely? Did the fires affect survival of P.
ramorum, and if so, what biological and environmental factors were most influential in
determining whether the pathogen survived the fires?
As part of our study examining the impact of wildfire on P. ramorum survival, we monitored
12 Big Sur watersheds for the pathogen in the spring of 2009. Stream monitoring, consisting
of both baiting with rhododendron leaves and vacuum water filtration through 3 μM
membranes, is a well-established method to assess P. ramorum presence on the landscape
scale. Baiting provides information on cumulative pathogen presence over several weeks,
while filtration allows for quantification of pathogen spores at a distinct time. Of the 12
watersheds, seven were burned in varying degrees and five were unburned. All of the
watersheds except one (Big Creek) were known to contain P. ramorum before the wildfires.
We were also interested to see if other Phytophthora species were present in different levels
between burned and unburned watersheds.
Using the baiting technique, P. ramorum was recovered from five of the seven burned
watersheds and four of the five unburned watersheds (Big Creek remained P. ramorum-free).
There were no obvious temporal trends in P. ramorum abundance over the four bait
deployment periods. From water filtration, over 250 isolates were subcultured and grouped
into 25 morphotypes; only one isolate was identified as P. ramorum. Roughly half of the total
isolates were suspected to be Phytophthora species, and these seemed to be equally
distributed between burned and unburned watersheds. DNA sequencing of the ITS1-ITS4
region identified several of the filtration isolates as P. gonapodyides and P. cactorum. Future
work includes additional DNA sequencing of water filtration isolates, as well as investigating
P. ramorum survival in previously infested, burned terrestrial plots.
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
Department of Plant Pathology, University of California, Davis 95616.
Corresponding author: mmbeh@ucdavis.edu.
292
Proceedings of the Sudden Oak Death Fourth Science Symposium
Multiplex Real-Time PCR for Detection of
Phytophthora ramorum, the Causal Agent
of Sudden Oak Death1
Guillaume J. Bilodeau,2 Richard C. Hamelin,3 Gervais Pelletier,3
Françoise Pelletier,3 and C. André Lévesque4
Abstract
Various molecular assays have been developed over the past few years for diagnostics of
Phytophthora ramorum. The redundancy obtained by using multiple gene regions has
increased the reliability in detecting the pathogen (Martin and others 2009). However, multigene assays require different PCR reactions to test a single sample. Having a reliable,
sensitive, specific, fast, and low-cost molecular diagnostic assay is highly desirable for P.
ramorum. In order to improve reliability and specificity without compromising speed and
cost, we propose the development of two multiplex real-time PCR assays (Bilodeau and
others 2009) using TaqMan probes with different reporter dyes targeting:
I) Phytophthora. ramorum and Phytophthora genus multiplex assay
Three different gene regions of Phytophthora ramorum (ITS, β-tubulin, elicitin) (Bilodeau
and others 2007), Phytophthora genus (β-tubulin).
II) Hierarchical and RuBisCO multiplexing
Phytophthora ramorum (ITS), Phytophthora genus (β-tubulin), oomycetes (ribosomal 5.8S
subunit) and host plants (RuBisCO), allowing simultaneous detection of P. ramorum, while
verifying DNA extraction and the presence of other oomycetes in the DNA sample. The
sensitivity of TaqMan assays in single and multiplex reactions was also compared in this
study.
These assays were tested on different Phytophthora species and oomycetes, and were verified
on two different sets of field samples previously assayed by other laboratories. These were
obtained from multiple field hosts infected by various Phytophthora species and the DNA
from one set was extracted from ELISA lysates.
When we used the multiplex assay combining three P. ramorum-specific TaqMan and one
Phytophthora genus TaqMan, all samples containing P. ramorum were detected and no false
negatives were obtained. However, some Phytophthora species cross-reacted with some
probes. Nevertheless, in all cases, a false positive with one probe was accompanied by
negative results for the other two P. ramorum-specific probes. This highlights the advantage
of the redundancy generated in a multiplex reaction.
All P. ramorum samples from pure cultures or field samples were detected using these
multiplex real-time PCR assays. In general, TaqMan multiplex assays showed lower detection
sensitivity than single separate reactions and, in some cases, lower fluorescence. However, the
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium,
June 15-18, 2009, Santa Cruz, California.
1
2
U.S. Department of Agriculture-Agricultural Research Service (USDA-ARS), Salinas, CA.
Natural Resources Canada, Canadian Forest Service, Laurentian Forestry Centre, Québec, Canada.
4
Agriculture and Agri-Food Canada, Central Experimental Farm - Biodiversity, Ottawa, Canada.
Corresponding author: Guillaume.Bilodeau@ars.usda.gov.
3
293
GENERAL TECHNICAL REPORT PSW-GTR-229
multiplex assays still detected all P. ramorum accurately while decreasing the cost and
increasing throughput.
The two multiplex assays developed here serve different purposes. The presence of an internal
control (the plant TaqMan probe) should reduce the rate of false negatives by identifying
reactions that have failed due to poor DNA extraction or to the presence of inhibitors. The
different probes developed here should be useful for diagnostics of plant pathogens and have
broader applications than just for P. ramorum. In plant health inspections or surveys, these
probes could serve as detection tools. In particular, the Phytophthora spp., oomycetes and
RuBisCO probes could be combined with other species probes for the detection of other plant
pathogens. We expect that these multiplex PCR assays will be useful in the direct detection of
P. ramorum from plant samples or ELISA solutions, as well as from cultures.
Literature Cited
Bilodeau, G.J.; Lévesque, C.A.; de Cock, A.W.A.M.; Duchaine, C.; Brière, S.; Uribe, P.;
Martin, F.N. and Hamelin, R.C. 2007. Molecular detection of Phytophthora ramorum by
real-time polymerase chain reaction using TaqMan, SYBR Green, and molecular beacons.
Phytopathology. 97: 632–642.
Bilodeau, G.J.; Pelletier, G.; Pelletier, F.; Lévesque, C.A. and Hamelin, R.C. 2009.
Multiplex real-time polymerase chain reaction (PCR) for detection of Phytophthora ramorum,
the causal agent of Sudden Oak Death. Canadian Journal of Plant Pathology. 31: 195–210.
Martin, F.N.; Coffey, M.D.; Zeller, K.; Hamelin, R.C.; Tooley, P.; Garbelotto, M.;
Hughes, K.J.D.; Kubisiak, T.; Bilodeau, G.J.; Levy, L.; Blomquist C. and Berger, P.
2009. Evaluation of molecular markers for Phytophthora ramorum detection and
identification: testing for specificity using a standardized library of isolates. Phytopathology.
99: 390–403.
294
Proceedings of the Sudden Oak Death Fourth Science Symposium
Within Tree Intensive Sampling Leads to the
Recovery of Multiple Genotypes of
Phytophthora ramorum in Southern Oregon
Tanoaks1
Jennifer Britt2 and Everett Hansen2
Abstract
Phytophthora ramorum can infect leaves, stems, or the bark of plants; persists in streams and
soils; and is even found in the xylem of oaks. Yet how P. ramorum infects and spreads
through individual trees and spreads through forests is not completely understood. Individual
trees often have multiple lesions on twigs in the crown, along branches, and on the main bole.
Sometimes the lesions appear to be connected in the bark or wood and sometimes not. Several
modes of infection are currently under investigation. These include mycelial spread within the
tree, rain splash dispersal, and wind dispersal resulting in an initial infection in the tree
canopy. In order to better understand how P. ramorum spreads through forests and individual
trees or plants, we are using DNA fingerprinting (microsatellite markers) to track the
movement of P. ramorum infections within trees in southern Oregon tanoak forests.
We isolated tissue from multiple crown and bole lesions of tanoak trees from 10 sites in
southern Oregon. We collected samples from 2001 to 2007 in a variety of ways. For the
within trees studies, trees were felled and the bark was scraped back to reveal lesions. In the
crown studies, trees were felled and mostly stems were sampled. All other samples were
collected during routine monitoring of southern Oregon forests by Oregon Department of
Forestry personnel. Samples were plated onto Phytophthora selective CARP medium in the
field and/or lab and grown out in the lab for identification (Hansen and others 2005). Isolates
were genotyped at five microsatellite loci PrMS39, PrMS43, and PrMS45 (Prospero and
others 2004), and PrM82 (Ivors and others 2006). Alleles were sized on an ABI Prism 3100
sequencer and results were analyzed using GeneScan and Genotyper software (Applied
Biosystems). Approximately 80 percent of isolates were re-extracted and/or re-genotyped to
confirm allele sizes.
A total of 54 trees from 2001 to 2007 were sampled at 10 sites in southern Oregon. Of those,
46 percent of tanoaks sampled had one genotype (25 trees) and 54 percent had greater then
one genotype (29 trees), including one tree with eight different genotypes.
Mapping multilocus genotypes onto individual trees demonstrates multiple infections of P.
ramorum by multiple individual zoospores in an individual plant are common. These results
also provide insight into modes of P. ramorum infection. In about 25 percent of the trees with
multiple genotypes there is vertical disconnection among genotypes where genotypes found at
the top of the tree are different then at the base. This suggests the base of the tree is infected
by rain splash versus the mid trunk which is likely infected by zoospores running down the
trunk during heavy rains. Multiple genotypes found in the crown suggest a mode of infection
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
Department of Botany and Plant Pathology, Oregon State University, Corvallis, OR 97331.
Corresponding author: brittje@science.oregonstate.edu.
295
GENERAL TECHNICAL REPORT PSW-GTR-229
via wind dispersal or rain splash by different P. ramorum individuals in the canopy. Once an
infection is established, our results also suggest the pathogen can move through the wood via
mycelial growth where one genotype is present and the lesions are connected. It is also
possible that a single genotype from multiple individual zoospores can inflict multiple
infections on an individual tree, where one genotype is dominant and clustered in three
separate areas.
Alternatively, it is possible the microsatellites are mutating as they move through a tree. This
would suggest the multilocus genotypes we found are not necessarily due to multiple
infections, but are actually mutation in action. This is unlikely because many of the genotypes
differ at more then one allele. Such a high rate of mutation has not, to our knowledge, been
demonstrated in any organism. We are currently analyzing the results of a plating and
sproutlet inoculation study where we genotyped individual isolates before inoculation and
after growth to see if we can pick up a mutation signal in order to suggest the actual mutation
rate of P. ramorum.
This work provides insight into how P. ramorum moves through individual plants postinfection and how an infection can move through a forest or nursery. It will hopefully help
managers trying to track the spread of sudden oak death through a forest or across continents
by molecular techniques. It also suggests that individual samples from a forest or nursery may
not tell the complete migration story and care should be taken to collect a representative
sample of isolates even within individual plants.
Acknowledgments
We would like to thank the U.S. Department of Agriculture, Forest Service, Pacific Southwest
Research Station, for funding; Alan Kanaskie and the Oregon Department of Forestry for
isolate collection and tree falling; and Paul Reeser, Jennifer Parke, Ebba Peterson, Wendy
Sutton, and Sonny Lucas for lab and field help.
Literature Cited
Ivors, K.L.; Garbelotto, M.; De Vries, I.; Ruyter-Spira. C.; Te Hekkert, B.; Rosenzweig,
N. and Bonants, P. 2006. Microsatellite markers identify three lineages of Phytophthora
ramorum in US nurseries, yet single lineages in US forest and European nursery populations.
Molecular Ecology. 15: 1493–1505.
Prospero, S.; Black, J.A. and Winton, L.M. 2004. Isolation and characterization of
microsatellite markers in Phytophthora ramorum, the causal agent of sudden oak death.
Molecular Ecology Notes. 4: 672–674.
Hansen, E.M.; Parke, J.L. and Sutton, W. 2005. Susceptibility of Oregon forest trees and
shrubs to Phytophthora ramorum: a comparison of artificial inoculation and natural infection.
Plant Disease. 89: 63–70.
296
Proceedings of the Sudden Oak Death Fourth Science Symposium
Waiting for SOD: Sudden Oak Death and
Redwood National and State Parks1
Monica Bueno,2 Janelle Deshais,2 and Leonel Arguello2
Background
Phytophthora ramorum is a highly aggressive, exotic pathogen responsible for the
deaths of several million oak (Quercus spp.) and tanoak (Lithocarpus densiflorus)
trees in California and Oregon and for leaf blight and twig die-back on over 100 other
native and ornamental plants (Rizzo 2008). Phytophthora. ramorum and the disease it
triggers, sudden oak death (SOD), has the potential to cause extensive ecosystem
changes in the forests of Redwood National and State Parks (RNSP). The climate and
forest systems of the north coast of California place RNSP in the highest risk
category for SOD infection (Meentemeyer and others 2004). Although to date it has
not been found in RNSP, the pathogen is present in forests 17 km north of the parks
in Curry County, Oregon and in streams 15 km south of the parks in McKinleyville,
California.
Redwood National and State Parks hold approximately 45 percent of the last
protected old-growth redwood forests left in California as well as many thousands of
acres of second growth forests. Tanoak is a major ecological component of these
forests and is proving to be especially susceptible to P. ramorum. Tanoak populations
infected by P. ramorum are approaching 100 percent mortality in some areas
(Meentemeyer and others 2008, Moritz and others 2008) and the potential for local
extinctions of tanoak is becoming more probable (Hansen 2008). Although the
specific ecological consequences related to the loss of tanoaks in RNSP forests are
unknown, we can assume there will be significant ecosystem changes due to the
keystone/foundation species status of tanoak (Ellison and others 2005). In the forests
of RNSP, tanoak is by far the principal mast producing species and provides foraging
and nesting substrate for a variety of wildlife both on the tree itself and in leaf litter
(Raphael 1986). RNSP also has an important cultural legacy of large stands of old
tanoak trees that have been managed by Native American families for many
generations.
Redwood National and State Parks has the unique opportunity to draw from over 10
years of SOD research and the management attempts of other agencies to develop a
SOD management plan before the disease arrives in the park. It is the goal of the
natural resources managers of RNSP to slow the inevitable arrival of P. ramorum to
the parks by implementing preventive measures; to prepare for its arrival by
conducting early detection for the pathogen and modeling the potential niche of its
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
Redwood National and State Parks, 121200 Highway 101, Orick, CA.
Corresponding author: monica_bueno@nps.gov.
297
GENERAL TECHNICAL REPORT PSW-GTR-229
most important hosts; and to address the actions needed to confront the disease once
it arrives by creating a SOD management plan based on ecological principles, the
basic tenets of disease management, a strong knowledge of our forest systems, and
early collaboration and consultation with regulatory agencies and adjacent land
managers. It is our hope that through this process, the park can be prepared to act
quickly and decisively to manage this pathogen when it finally does arrive.
Strategic planning
The first phase of our strategic plan consists of actions we are currently engaged in:
prevention, early detection, and management planning. Our management plan will
consider all tools available to combat this disease including strategies to eradicate,
contain, conserve, and restore where possible (table 1). Our second phase will begin
once the pathogen has arrived in the parks and we begin to implement the treatments
outlined in the management plan.
Prevention
Our education and outreach program is geared toward informing RNSP staff and
visitors about the seriousness of the disease and what they can do to help. We created
best management practices brochures for employees, contractors, and researchers
who work in RNSP and placed SOD information signs at visitor centers and
trailheads. We also have a SOD link on the public National Park Service (NPS)
website for RNSP.
Our prevention program also includes sanitation. We identified areas at high risk of
exposure to SOD by human vectors, such as muddy areas near high densities of
California bay laurels (Umbellularia californica) and tanoaks that are frequented by
park visitors. In an attempt to reduce this potential risk we are graveling these areas
to limit standing water and mud. This action is also helpful in our endeavor to reduce
the spread of Port-Orford-cedar root disease caused by Phytophthora lateralis. Other
prevention and sanitation measures, such as trail closures and cleaning stations, are
being considered.
Early Detection
Early detection and a quick response are the keys to successful management of SOD.
The RNSP early detection program includes stream baiting surveys, ground surveys,
and aerial surveys (flown by the U.S. Department of Agriculture, Forest Service and
ground checked by RNSP employees). We are also using an ecological niche model
of host species to guide early detection efforts and direct management decisions once
the disease has arrived.
Management Planning
Internal discussions are underway to understand the full array of options available to
park staff to manage this disease once it arrives. Options being considered include:
no action, preventive options, a containment strategy, an eradication strategy and
forest restoration, in both the short- and long-term time frames. Important
considerations in these discussions include funding, regulatory responsibilities
(natural and cultural consultations), adjacent landowner actions and responsibilities,
and public support. As part of our planning we may consider creating tanoak refuges
298
Proceedings of the Sudden Oak Death Fourth Science Symposium
(defined as tanoak groves that are least likely to become infected due to spatial or
temporal factors) and protecting them through the creation of no-host buffers. It is
our goal that through this planning process RNSP will be able to more effectively and
quickly implement the most appropriate treatment strategy at an initial infection site.
Sudden oak death management at RNSP will not be possible unless the treatment
options are consistent with the overall goals for resource management at the parks
and which take into consideration the high public profile position of the parks. The
RNSP mission is, in part, to “preserve significant examples of the primeval coastal
redwood forests and the prairies, streams, seashore, and woodlands with which they
are associated...” as well as to maintain park forests “in a condition of unimpaired
ecological integrity” (United States Department of the Interior 2000). We anticipate
there will be much discussion and debate about how to best manage this disease and
protect our forested ecosystems, consistent with park mission and values.
Table 1—Sudden Oak Death Strategic Planning Efforts at Redwood National
and State Parks
Strategy
Action
Goal
Current efforts (P. ramorum not present)
Prevention
Education/Outreach
BMPs for RNSP
Information signs
Public webpage
Sanitation
Gravel high-risk areas
To keep P.
ramorum out of
RNSP as long
as possible.
Early Detection
Stream baiting
Ground surveys
Aerial surveys (USDA Forest Service)
Host modeling
Management Planning
Communicate with regulatory agencies,
adjacent land owners, and interested public
on the full array of options available for
managing this disease consistent with park
values and authority.
To find the
pathogen
before it
becomes
established in
the landscape.
To develop and
implement a
sound
management
strategy to
combat this
disease in the
park.
Future potential efforts after P. ramorum arrives in park
Eradication
Set quarantine area
Some combination of herbicide, cut, burn
Monitor treated sites
To eliminate
pathogen
before it
spreads to
uninfected
sites.
299
GENERAL TECHNICAL REPORT PSW-GTR-229
Protection/Containment
Chemical treatments
AGRI-FOS®* – tanoaks
Physical treatments
Host removal in potential spread areas
(buffers)
Trail closures
Cleaning stations
Restoration/Conservation
Plant resistant hosts (if available)
Plant conifers and other species
Identify tanoak refuges and protect through
physical and chemical buffers
To strengthen
important trees
near infection
sites and to
prevent human
and natural
transmission of
the pathogen.
To conserve
ecologically
important oaks
and tanoaks in
the landscape,
rehabilitate
infection sites,
and minimize
ecological
impacts.
*Systemic fungicide approved for control and prevention of P. ramorum in California.
AGRICHEM Mfg. Ind. Pty. Ltd.
Literature Cited
Ellison, A.M.; Bank, M.S.; Clinton, B.D. [and others]. 2005. Loss of foundation species:
consequences for the structure and dynamics of forested ecosystems. Frontiers in Ecology and
the Environment. 3: 479–486.
Hansen, E.M. 2008. Rethinking Phytophthora - Research opportunities and management. In:
Frankel, S.; Kliejunas, J.T.; Palmieri, K.M., tech. coords. Proceedings of the sudden oak death
third science symposium. Gen. Tech. Rep. PSW-GTR-214. Albany, CA. U.S. Department of
Agriculture, Forest Service, Pacific Southwest Research Station: 5–14.
Meentemeyer, R.; Rank, N.; Shoemaker, D. [and others]. 2008. Impact of sudden oak
death on tree mortality in the Big Sur ecoregion of California. Biological Invasions. 10: 1243–
1255.
Meentemeyer, R.; Rizzo, D.; Mark, W. and Lotz, E. 2004. Mapping the risk of
establishment and spread of sudden oak death in California. Forest Ecology and Management.
200: 195–214.
Moritz, M.A.; Moody, T.; Ramage, B. and Forestel, A. 2008. Spatial distribution and
impacts of Phytophthora ramorum and sudden oak death in Point Reyes National Seashore.
California Cooperative Ecosystems Study Unit Task Agreement No. J8C07050015.
http://www.nps.gov/pore/naturescience/upload/diseases_sod_in_pore_080207.pdf. (3 June
2009).
Raphael, M.G. 1986. Wildlife-tanoak associations in Douglas-fir forests of northwestern
California. In: Plumb, T.R.; Pillsbury, H.H., eds. Symposium on the multiple-use
management of California's hardwood resources. Gen. Tech. Rep. PSW-100. San Luis
Obispo, CA: U.S. Department of Agriculture, Forest Service, Pacific Southwest Forest and
Range Experiment Station, 183–189.
Rizzo, D. 2008. Phytophthora ramorum: A recent discovery with a large impact.
Phytopathology. 98: S197.
300
Proceedings of the Sudden Oak Death Fourth Science Symposium
United States Department of the Interior. 2000. General management plan/General plan
Redwood National and State Parks, Humboldt and Del Norte Counties, California. Denver,
CO: National Park Service. [Pages unknown].
301
GENERAL TECHNICAL REPORT PSW-GTR-229
Effect of Fungicides on the Isolation of
Phytophthora ramorum from Symptomatic
and Asymptomatic Rhododendron
Leaf Tissue1
Gary Chastagner,2 Annie DeBauw,2 and Kathy Riley2
Abstract
A number of systemic and contact fungicides, such as Subdue MAXX, Dithane, and Maneb,
have been shown to be effective in controlling Phytophthora ramorum development on
several nursery crops (Chastagner and others 2006, Garbelotto and others 2002, Heungens
and others 2006, Linderman and Davis 2006, Tjosvold and others 2008). Studies on
rhododendrons have also shown that some fungicides, such as Captan, have very limited
residual activity, while residues of other fungicides, such as Segway, can significantly reduce
disease development up to 92 days after application (Chastagner and others 2009). The use of
fungicides to control P. ramorum has raised concerns that fungicide residues may affect the
ability to isolate this pathogen from infected tissue, or more importantly, mask symptom
development. Concealment of symptoms may make it more difficult to detect infected plants
during routine visual inspections.
As part of our ongoing work relating to the management of P. ramorum on nursery stock, we
conducted a study to determine if fungicide residues affected isolation of the pathogen from
symptomatic tissue, and if any of the fungicides suppressed symptom development. Thirteen
fungicides were included in this test (table 1).
The foliage on five container-grown ‘Nova Zembla’ rhododendron plants was sprayed with
each fungicide during late August. Two days later, three leaves were removed from each plant
and inoculated with suspensions of zoospores from an NA1 lineage rhododendron isolate of
P. ramorum. A total of six 10 ul drops of suspension were pipetted onto the lower leaf
surface, three on each side of the leaf midrib. The leaf tissue beneath three drops on one side
of the leaf midrib was injured using an insect pin, while the tissue beneath the drops on the
other side of the leaf was left unwounded. Checks included inoculated and non-inoculated
leaves from untreated plants.
After 7 days incubation at 19 to 20 ºC, the leaves were photographed and the resulting leaf
spots were measured using the APS ASSESS program. To evaluate the possibility of adverse
fungicide effects on isolation of P. ramorum from symptomatic and asymptomatic inoculation
sites, leaves were surface-sterilized in a 1:9 dilute bleach solution for 30 seconds prior to
plating tissues from all of the inoculation sites onto CARP selective medium.
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium,
June 15-18, 2009, Santa Cruz, California.
2
Washington State University, Puyallup Research and Extension Center, 2606 W. Pioneer,
Puyallup, WA 98371.
Corresponding author: chastag@wsu.edu.
302
Proceedings of the Sudden Oak Death Fourth Science Symposium
Table 1—Fungicides Included in Test
Product
Active Ingredient
Captan 80 WP
captan
Disarm 480 SC (TM437)
fluoxastrobin
Dithane 75 DF
mancozeb
Gavel 75 DF
mancozeb + zoxamide
Heritage WG 50
azoxystrobin
Insignia 20W
pyraclostrobin
Maneb 75 DF
maneb
NOA 446510/A12946
mandipropamid
Polyram 80 DF
metiram
Ranman 400SC (Segway)
cyazofamid
Stature DM 50 WP
dimethomorph
Subdue MAXX FV
mefonaxam
V-10161 4FL
fluopicolide
Isolations from fungicide-treated leaves indicated that none of the fungicides tested had any
adverse effect on the recovery of the P. ramorum from symptomatic tissue. P. ramorum was
recovered from 98.9 percent of the 933 symptomatic inoculation sites. However, isolations
from asymptomatic tissue suggest that Subdue MAXX and Insignia fungicides may pose a
high risk of masking symptom development on rhododendron. P. ramorum was recovered
from 12.1 percent of the 947 asymptomatic inoculation sites. About 67 percent of the
asymptomatic inoculation sites that yielded P. ramorum were on leaves that had been treated
with Subdue MAXX. About 17 percent were from leaves that had been treated with Insignia.
If additional trials confirm these results on rhododendron and other hosts, growers could
increase the effectiveness of visual inspections by using fungicides that do not adversely
affect symptom development.
Acknowledgments
This research was financially supported by the IR-4 Program and the Washington State
Department of Agriculture Nursery Research Program. The assistance of Jan Sittnick and Don
Sherry is gratefully acknowledged.
Literature Cited
Chastagner, G.A.; DeBauw, A.; Riley, K. and Dart, N.L. 2009. Residual effectiveness of
fungicides in protecting rhododendron leaves from Phytophthora ramorum. Phytopathology.
99: S180.
Chastagner, G.A.; Hansen, E.M.; Riley, K.L. and Sutton, W. 2006. Effectiveness of
fungicides in protecting Douglas-fir shoots from infection by Phytophthora ramorum. In:
Frankel, S.J.; Shea, P.J. and Haverty, M.I., tech. coords. Proceedings of the sudden oak death
second science symposium: the state of our knowledge. Gen. Tech. Rep. PSW-GTR-196.
Albany, CA: U.S. Department of Agriculture, Forest Service. 239.
Garbelotto, M.; Rizzo, D.M. and Marais, L. 2002. Phytophthora ramorum and sudden oak
death in California: IV. Preliminary studies on chemical control. In: Standiford, R.B.;
303
GENERAL TECHNICAL REPORT PSW-GTR-229
McCreary, D. and Purcell, K.L., tech. coords. Gen. Tech. Rep. PSW-GTR-184, Albany, CA:
U.S. Department of Agriculture, Forest Service. 811–818.
Heungens, K.; De Dobbelaere, I. and Maes, M. 2006. Fungicide control of Phytophthora
ramorum on rhododendron. In: Frankel, S.J.; Shea, P.J. and Haverty, M.I., tech. cords.
Proceedings of the sudden oak death second science symposium: the state of our knowledge.
Gen. Tech. Rep. PSW-GTR-196. Albany, CA: U.S. Department of Agriculture, Forest
Service. 241–257.
Linderman, R.G. and Davis, E.A. 2006. Evaluation of chemical and biological agents for
control of Phytophthora species on intact plants or detached leaves of rhododendron and lilac.
In: Frankel, S.J.; Shea, P.J. and Haverty, M.I., tech. cords. Proceedings of the sudden oak
death second science symposium: the state of our knowledge. Gen. Tech. Rep. PSW-GTR196. Albany, CA: U.S. Department of Agriculture, Forest Service. 265–268.
Tjosvold, S.A.; Koike, S.T. and Chambers, D.L. 2008. Evaluation of fungicides for the
control of Phytophthora ramorum infecting Rhododendron, Camellia, Pieris, and Viburnum.
Plant Health Progress. DOI: 10.1094/PHP-2008-0208-01-RS.
304
Proceedings of the Sudden Oak Death Fourth Science Symposium
Effect of Surface Sterilization Treatments on
the Detection and Viability of Phytophthora
ramorum on Various Substrates1
Katie Coats,2 Kathy Riley,2 Gary Chastagner,2 and Marianne
Elliott2
Abstract
An accurate evaluation of asymptomatic colonization of plant tissue by Phytophthora
ramorum requires the ability to distinguish between the surface contamination or epiphytic
growth of the pathogen and the colonization of plant tissues. Growth on a selective medium,
such as CARP, following the surface sterilization of plant tissue is often used to confirm P.
ramorum colonization of the tissue. Isolations and the use of PCR to detect asymptomatic
colonization require that epiphytic pathogen propagules and residual pathogen DNA are
rendered undetectable. A series of surface sterilization tests were performed with two
commonly used laboratory surface sterilants to determine their efficacy in removing epiphytic
propagules and DNA of P. ramorum in several different experimental scenarios. Substrates
tested include detached rhododendron (Rhododendron sp.) leaves, rhododendron leaf discs,
and freshly harvested Douglas-fir (Pseudotsuga menziesii) wood. Whatman filter paper was
included to represent a non-infectable substrate.
Results indicate that the efficacy of a treatment varies by experimental scenario and detection
method.
Rhododendron leaves and leaf discs: Based on post-sterilization growth on CARP medium, a
30-second treatment in a 10 percent solution of household bleach (0.6 percent sodium
hypochlorite) 1 hour after a zoospore suspension of P. ramorum was applied to rhododendron
leaves and leaf discs was as effective as higher concentrations of bleach or longer treatments
in bleach in killing the pathogen on the surface of this host. A reduced effectiveness of
sterilization treatments after a 3-hour post-inoculation incubation suggests that infection had
already occurred by the time of treatment.
Filter paper: To investigate the ability of sterilization treatments to eliminate epiphytic
propagules and residual pathogen DNA from the surface of a non-infectable substrate, discs
of Whatman filter paper were inoculated with a P. ramorum zoospore suspension, treated with
various sterilization treatments, and evaluated by quantitative PCR. Removal of pathogen
DNA with a 30-second treatment with 10 percent household bleach was as effective as higher
concentrations or longer exposure times. Neither 95 percent ethanol nor water was effective
at removing detectable P. ramorum DNA.
Douglas-fir wood chips and filter paper: Douglas-fir wood chips and filter paper discs were
inoculated with a P. ramorum zoospore suspension, incubated for 2 days, treated with various
surface sterilants, and then evaluated for presence of viable P. ramorum (isolation) and DNA
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
Washington State University, Puyallup Research and Extension Center, 2606 W. Pioneer, Puyallup,
WA 98371.
Corresponding author: kpcoats@wsu.edu.
305
GENERAL TECHNICAL REPORT PSW-GTR-229
(qPCR). A 30-second treatment of 10 percent household bleach, 95 percent ethanol, or water
was effective at eliminating detection of viable epiphytic P. ramorum on filter paper by
isolation, but only the bleach treatment eliminated detection of P. ramorum DNA by qPCR.
None of the treatments affected the detection of the pathogen by isolation or qPCR that had
internally colonized the Douglas-fir wood chips.
306
Proceedings of the Sudden Oak Death Fourth Science Symposium
Phosphonate Treatment to Control
Phytophthora cinnamomi Infection of Ione
Manzanita, A. myrtifolia1
Ellen Crocker2 and Matteo Garbelotto3
Abstract
We are investigating the use of the phosphonate Agri-Fos® as a preventative treatment for
Phytophthora cinnamomi infection of Arctostaphylos myrtifolia, Ione manzanita. A.
myrtifolia is a rare, federally threatened species endemic to the acidic, iron-oxide clay soils of
the Sierra Nevada foothills.
Habitat changes from mining, development, and changes in fire frequency have contributed to
its decline, and in recent years P. cinnamomi root and crown rot has further increased
mortality. P. cinnamomi is a widespread pathogen that attacks forest, ornamental, and
agricultural plants, and while phosphonate is commonly used to control P. cinnamomi
infections in general, its specific use for A. myrtifolia is unknown. The primary goal of our
study is to develop a treatment regime for A. myrtifolia, first determining the optimal: 1)
concentration of phosphonate/surfactant solution and 2) time of year for application.
To answer these questions we treated sample plots of A. myrtifolia with the following
surfactant solutions in the fall and spring: 0.005x Agri-Fos® without surfactant, 0.025x AgriFos® without surfactant, 0.05x Agri-Fos® without surfactant, 0.005x Agri-Fos® with
surfactant, and 0.025x Agri-Fos® with surfactant. At regular intervals cuttings were taken
from each plot, as well as control non-sprayed cuttings, and inoculated with either P.
cinnamomi (P3232) or V8 agar control. Eleven days after inoculation lesions were measured.
One month after application, both the spring and fall treatments significantly reduced lesion
length in infected cuttings. All tested solutions, regardless of concentration and surfactant
presence, appeared equally effective. Seven months after application, the spring treatment
continued to significantly reduce lesion length in infected cuttings, but results from the fall
treatment were mixed. For the spring treatment, all tested solutions were still equally
effective; however, only four of the five fall treatments significantly reduced lesion length and
even those appeared less effective than their spring counterparts.
These results suggest that spring Agri-Fos® treatment may be a useful tool in preventing P.
cinnamomi infection of A. myrtifolia. Because all concentrations appeared equally effective,
even a 0.005x topical application would reduce infection while having a minor environmental
impact. While Agri-Fos® treatment seems promising, further studies are needed before any
recommendations can be made. Because this treatment is somewhat phytotoxic, it is important
to determine whether the benefit to the plants outweighs the damage caused. Future studies
should determine whether Agri-Fos® is a good option despite phytotoxicity and how
phytotoxicity can be reduced. Ideally, this improved understanding of treatments will enable
us to outline a landscape level approach to prevent P. cinnamomi infection of A. myrtifolia.
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium,
June 15-18, 2009, Santa Cruz, California.
2
Cornell University, 334 Plant Sciences Building, Ithaca, NY 14853.
3 UC Berkeley, 137 Mulford Hall, Berkeley, CA 94720.
Corresponding author: matteo@nature.berkeley.edu.
307
GENERAL TECHNICAL REPORT PSW-GTR-229
Predicting Sudden Oak Death in Redwood
National and State Parks: Ecological Niche
Modeling of Key Transmission Host Species
Using Maximum Entropy1
Janelle Deshais,2 John Stuart,2 Steven Steinberg,3 and Leonel
Arguello4
Introduction
Phytophthora ramorum, the cause of sudden oak death (SOD), is an aggressive
introduced plant pathogen that has caused the death of several million tanoak
(Lithocarpus densiflorus) and oak (Quercus spp.) trees in the Pacific Northwest of
the United States (Rizzo 2008). Primarily an airborne pathogen, P. ramorum spreads
from its hosts via wind-driven rain. Not all hosts are equal; in the Pacific Northwest,
tanoak and California bay laurel (Umbellularia californica) are most critical for
wildland pathogen transmission (Rizzo 2005), and in the United Kingdom,
rhododendron species (Rhododendron spp.) are spreading the pathogen (DEFRA,
2007).
To date, P. ramorum has not been detected in Redwood National and State Parks
(RNSP); however, the pathogen is rapidly approaching. The Curry County, Oregon
infestation is 17 km north of RNSP’s Jedediah Smith State Park, and the recent
detection of P. ramorum in Norton and Mill Creeks, McKinleyville (Humboldt
County, California) is less than 15 km southwest of RNSP’s southern border.
Furthermore, key disease transmission species, California bay laurel, tanoak, and
Pacific rhododendron (R. macrophyllum), are abundant throughout much of RNSP.
Previous disease spread models predict RNSP in the highest risk category for SOD
disease susceptibility (Meentemeyer and others 2004, Guo and others 2005, Kelly
and others 2007). These spread models, however, predict establishment and risk on a
scale larger than RNSP needs to aid management decisions. Also, vegetation data
used in existing models are primarily derived from remotely sensed data which can
have limited success deciphering understory species in old-growth redwood forests.
Our main objective in this study was to create detailed (≤ 30 m² resolution) species
distribution maps for California bay laurel, tanoak, and Pacific rhododendron
throughout RNSP. Once model inputs are refined to yield highest possible accuracy,
we will use resulting species distributions to create a local-scale spread model. Final
host distributions and our spread model will be used to understand P. ramorum’s
potential spread extent and lethality within RNPS’s forests, as well as to guide
management decisions in the event that P. ramorum does arrive.
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
Department of Forestry and Wildland Resources, Humboldt State University, Arcata, CA 95521.
3
Institute for Spatial Analysis, Humboldt State University, Arcata, CA 95521.
4
Redwood National and State Parks, 121200 Highway 101 S, Orick, CA 95555.
Corresponding author: janelle_deshais@nps.gov.
308
Proceedings of the Sudden Oak Death Fourth Science Symposium
Methods
All species distribution modeling was carried out using MaxEnt v3.3, a niche-based
modeling program (Phillips 2005, Phillips and others 2006). We chose MaxEnt based
on its ability to handle categorical data as well as its use of presence-only records.
MaxEnt uses deterministic algorithms to create probability distributions, with
maximum entropy, for each target species based on a given set of predictor variables
(Phillips and others 2006). The modeling process relates occurrence of a species to
the broadest range of predictor variables, while optimizing predictive power.
MaxEnt requires species occurrence records and raster-formatted predictor variables.
All occurrence records were compiled, post hoc, from 11 vegetation surveys that
yielded a total of 1,606 unique point locations containing the presence of at least one
of our target host species. We ran analysis with all available host presence points as
well as with a subset that excluded points without confirmed geographic coordinates.
There were a total of 1,188 occurrence records available for tanoak, 566 for Pacific
rhododendron, and 216 for California bay laurel. The subset yielded 948 records for
tanoak, 426 for Pacific rhododendron, and 187 for California bay laurel. For all
analyses, 50 percent of occurrence records were used for model training and the
remaining 50 percent were held aside for model validation.
We tested seven environmental data layers (table 1) in raster format, managed in
ArcGis 9.3 (ESRI 2008). To minimize effects of autocorrelation within our models,
we included bias files (using bias option in MaxEnt) that described relative sampling
intensity (per survey area) throughout the study extent.
For each species we
Table 1. Parameters tested in preliminary niche models predicting
presence of Umbellularia californica , Lithocarpus densiflorus , and tested eight unique
Rhododendron macrophyllum throughout Redwood National and State parameter
combinations. Each
Parks.
trial included aspect,
Raster resolution
degree slope, distance
10m²
30m²
from ocean, elevation,
Species records*
soil type, and
All available
vegetation type;
Subset (excluded records without known GPS point)
however, we also ran
Environmental data
each model in 10 m²
Aspect (transformed to 0-180° scale, reclassified into 8 classes)
and 30 m² resolutions,
Degree slope
with both sets of
Distance from ocean
species records, and
Elevation
Slope position (continuous percentage)
tested slope position
Slope position (5 classes)
in two different
Soil type
formats (table 1). We
Vegetation type
then chose the highest
*Different bias layer, depending on which set of species records used
performing
combination (based on largest area under curve (AUC) value – see below) and reran
the model with 10 replicates (using random seed, with replacement). In all analyses
we enabled the jackknifing option to assess relative importance of each predictor
variable used in the model.
309
GENERAL TECHNICAL REPORT PSW-GTR-229
Results
Our preliminary results were encouraging, with all models performing significantly
better than random [receiver operating characteristic (ROC) AUC values > 0.5] and p
< 0.0001]. California bay laurel was best predicted by MaxEnt, followed by Pacific
rhododendron and tanoak, with AUC values of 0.914, 0.743, and 0.701, respectively
(table 2). Models with AUC values above 0.7 are generally considered useful, while
values above 0.9 indicate high performance (Swets 1988, Elith and others 2006).
Each species was modeled using slightly different sets of parameters (table 3). Final
species presence/absence distributions were produced by converting the continuous
logistic outputs (probability of presence) generated by MaxEnt, to binary outputs
(presence/absence) using threshold values (fig. 1, table 2).
Figure 1—Presence of Phytophthora
rramorum critical transmission host
species in Redwood National and
State Parks, California. Species’
predicted presence is shown in black.
Of the total 54,699 ha study site, our
model predicted 3,361 ha (6 percent)
with presence of Umbellularia
californica, 21,804 ha (40 percent)
with Rhododendron macrophyllum,
25,880 ha (47 percent) with
Lithocarpus densiflorus, and 33,163
ha (61 percent) with at least one of
these species present.
Presence/absence predictions were
calculated using binary thresholds
listed in table 2. All modeling was
carried out in MaxEnt v3.3 (Phillips
and others 2006).
310
Proceedings of the Sudden Oak Death Fourth Science Symposium
Table 2. Mean accuracy of preliminary niche models generated in MaxEnt v3.3. Each mean is
calculated from ten replicates (with replacement) of the 'best' model design (largest AUC) for
each species.
Species
Umbellularia californica
AUC (STDV)
0.914 (0.019)
Sensitivity (STDV)
0.907 (0.020)
Binary Threshold
0.388*
p Value
Rhododendron macrophyllum
0.743 (0.017)
0.763 (0.083)
0.317**
<0.0001
Lithocarpus densiflorus
0.701 (0.011)
0.862 (0.056)
0.324**
<0.0001
<0.0001
*Logistic threshold value set where model sensitivity (true positives) equals specificity (false positives)
**Logistic threshold value set to maximize sensitivity plus specificity
Table 3. Relative mean contribution of the four most informative environmental
variables used in each model. All means were calculated from ten replicates (with
replacement) of the 'best' model (largest AUC) for each species.
Species
Parameter
Relative Contribution (%)
Umbellularia californica
(10m² grid resolution used)
Elevation
Distance from ocean
Soil type
Degree slope
Other
47.78
20.54
14.31
7.18
10.18
Rhododendron macrophyllum
(10m² grid resolution used)
Soil type
Distance from ocean
Elevation
Slope position
Other
33.11
21.48
12.08
11.03
22.30
Lithocarpus densiflorus
(30m² grid resolution used)
Soil type
Distance from ocean
Vegetation type
Elevation
Other
35.32
27.76
13.17
9.90
13.84
311
GENERAL TECHNICAL REPORT PSW-GTR-229
Literature Cited
DEFRA [Department of Environmental Food and Rural Affairs; Forestry Commission];
Horticultural Development Council. 2007. Understanding Phytophthora ramorum: Key
findings from UK research. London, UK. 6 p. http://www.pdir.dk/Admin/Public/DWS
Download.aspx?file=%2FFiles%2FFiler%2FGartneri%2FPlanter%2FPramorum_Key_Resear
ch_Findings.pdf. (December 19, 2008).
Elith, J.G.; Anderson, C.H.; Dudı´k, R.P.; Ferrier, M.; Guisan, S.; Hijmans, A.;
Huettmann, R.J.; Leathwick, F.; Lehmann, J.R.; Li, A.; Lohmann, J.; Loiselle, L.G.;
Manion, B.A.; Moritz, G.; Nakamura, C.; Nakazawa, M.; Overton, Y.; Peterson, J.;
Phillips, A.T.; Richardson, S.J.; Scachetti-Pereira, K.S.; Schapire, R.; Sobero´n, R.E.;
Williams, J.; Wisz, M.S. and Zimmermann, N.E. 2006. Novel methods improve prediction
of species’ distributions from occurrence data. Ecography. 29: 129–151.
ESRI. 2008. ArcGIS. 9.3.Redlands, CA.
Guo, Q.; Kelly, M. and Graham, C. 2005. Support vector machines for predicting
distribution of Sudden Oak Death in California. Ecological Modeling. 182: 75–90.
Kelly, M.; Guo, Q.; Liu, D. and Shaari, D. 2007. Modeling the risk of a new invasive forest
disease in the United States: an evaluation of five environmental niche models. Computers,
Environment and Urban Systems. 31: 689–710.
Meentemeyer, R.; Rizzo, D.; Mark, W. and Lotz, E. 2004. Mapping the risk of
establishment and spread of Sudden Oak Death in California. Forest Ecology and
Management. 200: 195–214.
Phillips, S.J. 2005. Maxent software for species distribution modeling.
http://www.cs.princeton.edu/~schapire/maxent/. (July 20, 2008).
Phillips, S.J.; Anderson, R.P. and Schapire, R.E. 2006. Maximum entropy modeling of
species geographic distributions. Ecological Modelling. 190: 231–259.
Rizzo, D. 2008. Phytophthora ramorum: A recent discovery with a large impact.
Phytopathology. 98: S197.
Rizzo, D.M.; Garbelotto, M. and Hansen, E.M. 2005. Phytophthora ramorum: integrative
research and management of an emerging pathogen in California and Oregon forests. Annual
Review of Phytopathology. 43: 309–335.
Swets, J.A. 1988. Measuring the accuracy of diagnostic systems. Science. 240: 1285–1293.
312
Proceedings of the Sudden Oak Death Fourth Science Symposium
Phytophthora ramorum - Pathogenic
Fitness Among the Three Clonal Lineages1
Clare R. Elliott,2 Virginia McDonald,2 and Niklaus J. Grünwald3
Abstract
The Oomycete pathogen Phytophthora ramorum causes sudden oak death on oak and
ramorum blight on a wide range of ornamental plants, causing severe economic losses to the
nursery industry. The U.S. population of P. ramorum consists of three distinct clonal lineages
referred to as NA1, NA2, and EU1. We tested the hypothesis of differences in fitness among
these three lineages through the infection of detached leaves and whole plants in wounded and
non-wounded inoculations of rhododendron, and also in in vitro experiments to assess growth
and sporulation.
In independent experiments, the fitness of isolates within lineages was determined using the
fitness components lesion area (LA), sporulation capacity (SC), incubation period (IPw), and
the area under the lesion expansion curve (AULEC) on wounded detached leaves of two
cultivars of rhododendron. Three isolates from each clonal lineage were tested in vitro. In the
non-wounded whole plant experiments, inoculation was achieved by dipping plants in a
zoospore suspension of 5,000 zoospores/ml. Incidence was measured by the number of
infections and the number of leaves infected, and severity was measured by the total lesion
area and average lesion area per leaf. Ten P. ramorum isolates from each clonal lineage were
studied in vitro to assess growth and sporulation at 10 °C and 20 °C dark incubation for 10
days using Petri plates of V8 100 agar and repeated twice.
Lesion area demonstrated significant differences among lineages in two out of three wounded
detached leaf experiments; however, SC, IPw, and AULEC showed no consistent significant
differences among lineages. The non-wounded whole plant dip inoculations showed a trend
towards a difference between the NA1 lineage and EU1 and NA2 (0.1 > P > 0.05), but
variability among isolates within lineages means that these slight differences are not
statistically significant. In one out of the two experiments on whole plants, significant
differences between isolates within lineages were observed (P < 0.026). Analysis of variance
on in vitro growth and sporulation of isolates of P. ramorum showed that at 10 °C there was a
significant difference in the growth of isolates among lineages (P < 0.001) and at 20 °C there
was a trend towards a difference in sporulation of isolates among lineages, but these
differences weren’t significant (P = 0.075).
Both in vivo studies and in vitro experiments all point towards slight fitness differences
between clonal lineages of P. ramorum. Lineage NA1 generally grew more slowly, sporulated
less, and infected at a lower frequency and severity than lineages NA2 and EU1. In many
cases there was no significant difference between isolates of NA2 and EU1. In the whole
plant, EU1 isolates demonstrated slightly higher incidence than NA2, but lower severity (not
significantly different from NA1). More often than not these differences were not significant
due to a high degree of variability among isolates within lineages. Variability was greater in
lineages NA1 and EU1 compared with NA2 in in vitro growth and sporulation experiments.
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
Department of Botany and Plant Pathology, Oregon State University, Corvallis, OR 97331
3
Horticultural Crops Research Laboratory, USDA ARS, Corvallis, OR 97330.
Corresponding author: nik.grunwald@ars.usda.gov.
313
GENERAL TECHNICAL REPORT PSW-GTR-229
Susceptibility, Severity, and Sporulation
Potential of Phytophthora ramorum on
Several Rhododendron Species and
Hybrids1
Marianne Elliott,2 Gary Chastagner,2 Katie Coats,2 Annie DeBauw,2
and Kathy Riley2
Introduction
Several plant genera are important in the spread of Phytophthora ramorum in
nurseries. These include Rhododendron, Camellia, Viburnum, Pieris, and Kalmia. In
an effort to reduce the risk of P. ramorum being introduced onto their sites, some
nurseries have reduced or eliminated these species from their inventory. Of these
genera, Rhododendron is responsible for many of the P. ramorum positive finds in
nurseries in the Pacific Northwest. With input from industry representatives,
Washington State Department of Agriculture (WSDA), and the Rhododendron
Species Foundation (RSF) we have identified approximately 100 Rhododendron
hybrids and species to include in our testing. Three criteria were used to select
species from the RSF collection for inclusion in this study: species that are native to
the Yunnan province of China (it is speculated that P. ramorum may have originated
from this area); species which are commonly cultivated; and species that are
commonly used in hybridizing.
Methods
Leaves from 42 Rhododendron species were collected in November 2008 and
February 2009 at the Rhododendron species garden, Federal Way, Washington
(WA). A collection of 58 Rhododendron hybrids identified as being high priority for
testing by the WA nursery industry was also sampled in November 2008. Treatments
consisted of wounded and non-wounded inoculations with 10 µl of a zoospore
suspension of P. ramorum (NA1 population) on the upper and lower leaf surfaces. A
set of wounded treatments with a 10 µl drop of sterile distilled water was included for
each species as a control. Leaves were placed on moist filter paper in Petri dishes and
incubated at 19 C in the dark. Photos were taken at 5 and 10 days post-inoculation.
Lesions were measured using APS ASSESS and lesion size expressed in mm2.
Ten 6 mm diameter disks were cut from lesioned areas of leaves. If there was no
lesion present, the area including the inoculum site was used. Disks were incubated in
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
Washington State University, Puyallup Research and Extension Center, 2606 W. Pioneer, Puyallup,
WA 98371.
Corresponding author: melliott2@wsu.edu.
314
Proceedings of the Sudden Oak Death Fourth Science Symposium
a Petri dish containing 1 ml sdH2O and covered with nylon mesh to break surface
tension. After 10 days, the disks were transferred to a 2 ml cryovial containing 1 ml
sdH2O and shaken on a tissue homogenizer for 40 s. The disks were removed from
the spore suspension and spores were counted in five 30 µl aliquots placed on a
microscope slide at 100 x. Both sporangia and chlamydospores were counted and
spore counts were expressed as spores/ml and spores/lesion. The lesion area resulting
from the wounded lower surface was used in these calculations.
Data for the fall 2008 and winter 2008 collections were analyzed using t-tests.
Rhododendron species and hybrids were grouped into with/without indumentum,
lepidote/elepidote, and species originating in Yunnan/other places. A correlation test
was done comparing the fall and winter 2008 data for Rhododendron species.
Results
The wounded treatment produced larger lesions than the unwounded treatment.
Infection frequency was also higher on the wounded treatment for both upper and
lower leaf surfaces. Lesions were larger on the lower surface in both wounded and
unwounded treatments, and infection frequency was higher on the lower surface.
Species with indumentum on the lower leaf surface were less susceptible and
produced smaller lesions than species without indumentum.
Rhododendron species and hybrids with scales on the lower leaf surface were more
susceptible to infection and produced bigger lesions than those without, but the
difference was not significant in the November 2008 sampling period for
Rhododendron species. Species originating from Yunnan Province, China, were less
susceptible to infection and had smaller lesions than those from other areas.
Sporulation was not influenced by presence of indumentum or scales, and there was
no significant difference in sporulation between species originating from Yunnan and
those from other places. There was a significant positive correlation between lesion
size and sporangia/ml (r = 0.509, P = 0.001). There was no significant correlation
between chlamydospores/ml and lesion size or sporangia/ml.
Preliminary data indicate that there is a great deal of variation among the
Rhododendron species in our tests in the amount of sporangia and chlamydospores
produced on the infected leaf discs. This may be related to leaf surface features such
as indumentum and scales, and possibly chemical differences. Extremely high
numbers of chlamydospores were produced on foliage of R. campanulatum (319/ml).
R. brachycarpum produced the most sporangia (289/ml), more than 5 times the
amount produced by R. ponticum (50/ml), the species responsible for spread of P.
ramorum from European urban gardens to urban forests. Species that had low
susceptibility to infection and low sporulation potential included R. arboreum and R.
keiskei. R. dauricum was one of the most susceptible species and had high sporangia
production (184/ml). Although additional testing is needed to confirm these results, it
is clear that some rhododendrons pose a much higher risk of spreading P. ramorum
than others.
315
GENERAL TECHNICAL REPORT PSW-GTR-229
Acknowledgments
The authors wish to thank the Rhododendron Species Foundation, Washington State
Department of Agriculture, Briggs Nursery, Dennis Bottemiller, and Dan Meier.
316
Proceedings of the Sudden Oak Death Fourth Science Symposium
Phenotypic Variation in Phytophthora
ramorum: Wild Type vs Non-Wild Type
Isolates1
Marianne Elliott,2 Grace Sumampong,3 Simon F. Shamoun,3 Elisa
Becker,3 Aniko Varga,4 Delano James,4 Saad Masri,4 and Niklaus
J. Grünwald5
Abstract
Phenotypic characteristics of four Phytophthora ramorum isolates with atypical culture
morphology (non-wild type; nwt) were compared with four “wild type” (wt) isolates using
material from stock cultures and after re-isolation from lesions on inoculated rhododendron
leaves. Our preliminary results show that nwt isolates were more variable than wt isolates in
all of the characters tested, and were generally lower in aggressiveness, chlamydospore
production, and growth rate at all temperatures for both the original culture and when reisolated from a host.
Introduction
In earlier studies, unusual culture morphology and behavior were noticed among
some NA1 isolates of Phytophthora ramorum. This “non-wild type” behavior was
not observed in our collection of isolates from the EU1 or NA2 lineages, even though
the isolates had been in culture for a similar amount of time. It has been suggested
that subculturing in vitro causes culture instability and loss of virulence, and passage
through the host can revive the isolate back to its original state. To study this, we
compared four less virulent isolates (non-wild type; nwt) with four isolates of normal
virulence (wild type; wt) in our culture collection. One objective of this study was to
determine whether wt behavior could be restored to nwt isolates of P. ramorum by
successive re-isolation from host material.
Methods
Eight isolates of P. ramorum were selected and maintained on 15 percent V8 agar.
Phenotypic characters examined on original cultures were pathogenic aggressiveness;
growth rate at maximum, optimum, and minimum temperatures; and chlamydospore
production in vitro. Detached leaves of Rhododendron “Cunningham’s White” were
inoculated with each of the isolates and lesion size measured using APS ASSESS,
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
Washington State University, Puyallup Research and Extension Center, Puyallup, WA 98371.
3
Canadian Forest Service, Pacific Forestry Centre, Victoria, BC, Canada V8Z 1M5.
4
Sidney Laboratory, Canadian Food Inspection Agency, Sidney, BC, Canada V8L 1H3.
5
Horticultural Crops Research Laboratory, USDA ARS, Corvallis, OR 97330.
Corresponding author: melliott2@wsu.edu.
317
GENERAL TECHNICAL REPORT PSW-GTR-229
and then P. ramorum was isolated from lesions onto PARP and transferred to 15
percent V8 agar. These re-isolates were inoculated onto rhododendron leaves and reisolated two more times, for a total of three successive re-isolations.
Growth rate at maximum, optimum, and minimum temperatures, and chlamydospore
production were measured on cultures from the original and first re-isolation for each
isolate.
Results
In both wt and nwt groups, there were significant differences in lesion size on
detached rhododendron leaves between the original culture and the first re-isolation.
Successive re-isolations were not different from the original culture and the first reisolation. After re-isolation from the host, nwt isolates were still less aggressive than
wt isolates. Along with lower aggressiveness on rhododendron leaves, nwt isolates
produced fewer chlamydospores in V8 agar than did wt isolates. There was no
difference in growth rate between the original culture and the first re-isolation for
most isolates. However, nwt isolates were found to be more sensitive to temperatures
below 2 C and above 28 C. The optimum growth temperature was 20 C for both
wt and nwt isolates.
Non-wild type isolates were more variable than wt in all characters tested. The
greater variability suggests that these isolates are unstable or that slightly deleterious
mutation(s) have accumulated in accordance with Muller’s ratchet resulting in
reduced fitness. Wt isolates performed better than nwt isolates in all of the phenotypic
characters examined. Why nwt survives and proliferates is still a mystery. To
understand the cause of these phenotypic differences, the role of cytoplasmic
elements and differences in mitochondrial and nuclear DNA are being examined.
Further studies will also include examining sporulation of wt and nwt isolates on
plant hosts.
Acknowledgments
The authors wish to thank the Natural Sciences and Engineering Research Council of Canada,
Canadian Forest Service, and the Canadian Food Inspection Agency for financial support.
Partial funding to NJG for this work was also provided by the USDA ARS CRIS project
5358-22000-034-00.
318
Proceedings of the Sudden Oak Death Fourth Science Symposium
Ecology of Phytophthora ramorum in
Watercourses: Implications for the Spread
and Management of Sudden Oak Death1
Elizabeth Fichtner,2 Kamyar Aram,2 and David Rizzo2
Abstract
Though stream-baiting has proven useful for early detection of new terrestrial infestations of
Phytophthora ramorum, the biological and ecological rationale behind the success of baiting
are unknown. Our current studies address the central theme of “why baiting works,” focusing
on specific questions including: i) What are the inoculum sources in streams? ii) Is there an
inoculum threshold for baiting? iii) Can zoospore cysts undergo diplenetism to infect bait
leaves? and iv) what are the trophic niches of P. ramorum in streams, and how does this
influence bait detection?
To address the potential for P. ramorum to persist on aquatic or riparian plants, several plant
species were collected from a P. ramorum-infested holding pond in Humboldt County,
California. Aboveground plant parts and roots were baited with rhododendron leaf disks;
however, thus far, P. ramorum was not recovered from any plant tissues. Additionally,
aquatic plants have been inoculated to determine their susceptibility to P. ramorum, but the
pathogen has not been re-isolated from any inoculated tissues. To assess the role of trophic
niche on bait detection of P. ramorum, living and dead rhododendron leaves were incubated
in two streams infested with the pathogen. P. ramorum colonized only living bait tissues,
suggesting that the pathogen serves as a biotroph with respect to baiting in streams.
Laboratory experiments were designed to address the inoculum threshold necessary for bait
detection and the potential for cysts to undergo diplanetism and subsequently infect baits. P.
ramorum cysts, ranging in concentration (0, 101, 102, 103, 104), were placed in 1.5 ml
microfuge tubes containing 1 ml of water, in the presence or absence of a surfactant. Cyst
suspensions were centrifuged and then leaf disk baits were incubated on the water surfaces for
48 hours. Leaf disk baits were never infected in the presence of a surfactant. In the absence of
surfactant, baits were successfully infected at cyst concentrations of 102 and 103 cysts/ml. The
disparity between bait infectivity in the presence/absence of surfactant suggests that cysts may
diplanetize, forming motile infective zoospores.
Preliminary results suggest that P. ramorum behaves as a biotroph when infecting bait
materials in streams; however, the ability of P. ramorum to infect and persist on aquatic plants
or colonize detritus in streams is unknown. Germination of cysts to form motile zoospores
may aid in bait detection by enabling pathogen homing, thus enhancing bait efficacy at low
inoculum concentrations.
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
University of California, Department of Plant Pathology, One Shields Ave., Davis, CA 95616.
Corresponding author: ejfichtner@ucdavis.edu.
319
GENERAL TECHNICAL REPORT PSW-GTR-229
Emergence of Phytophthora cinnamomi in a
Sudden Oak Death-Impacted Forest1
Elizabeth J. Fichtner,2 David M. Rizzo,2 Tedmund J. Swiecki,3 and
Elizabeth A. Bernhardt3
Abstract
The introduction of Phytophthora ramorum to China Camp State Park (CCSP) (Marin
County, California) in the 1990s resulted in extensive mortality of Quercus agrifolia by 2000.
However, mortality is now occurring in discrete disease centers among species of trees that
were not affected by the sudden oak death (SOD) epidemic. This new mortality was first
observed in long-term plots established by Phytosphere Research for study of SOD.
Symptoms observed on Pacific madrone (Arbutus menziesii) included wilting followed by
rapid mortality. In contrast, California bay laurel (Umbellularia californica) within affected
areas typically showed thinning of the canopy and foliar chlorosis, followed by progressive
top dieback. Most of the killed bay laurel observed to date have been saplings or small trees,
whereas madrone mortality includes larger trees. Symptoms were not congruent with those
associated with SOD, but were suggestive of a root disease.
Soil and root samples were collected beneath symptomatic trees. Roots were surface sterilized
and embedded in PARP, whereas soil samples were baited with rhododendron leaf disks. P.
cinnamomi was baited from multiple soil samples within the affected area and was isolated
from roots of symptomatic bay laurel and madrone. P. cambivora was baited from soils in one
portion of the affected area, but was not isolated from roots.
Containerized bay laurel and madrone were purchased for inoculation with P. cinnamomi and
completion of Koch’s Postulates. Pots of individual plants were placed in plastic bags to allow
for periodic saturation of the container soil mix. Six plants of each species were inoculated by
pouring a 15 ml suspension containing 105 zoospores/ml on the saturated soil surface; four
plants of each species served as uninoculated controls. Pots were flooded for 24 hours after
inoculation and then drained to container capacity. Pots were re-flooded for 24 hour intervals
once each week over a 3-week period. The flood water was baited for Phytophthora during
each flood event. After 3 weeks, roots were excavated from pots, surface sterilized, and baited
with rhododendron leaf disks.
Bay laurel and madrone wilted within 2 weeks of inoculation with P. cinnamomi. Roots of
inoculated plants were necrotic, and P. cinnamomi was re-isolated from symptomatic roots. In
the initial run of this experiment, one of the non-inoculated control madrones died; P.
cactorum was isolated from roots of this plant. In the second run of the experiment, P.
cinnamomi was isolated from foliar lesions on a non-inoculated madrone. In subsequent trials,
most non-inoculated plants remained asymptomatic; however, P. cinnamomi, P. cambivora,
P. cactorum, Pythium sterilum, and Pythium vexans were isolated from surface-sterilized
roots of non-inoculated container-grown madrone. P. syringae and P. cinnamomi were
isolated from symptomatic foliage and stems of non-inoculated madrone. Additionally, P.
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
Department of Plant Pathology, UC Davis, Davis, CA 95616.
3
Phytosphere Research, Vacaville, CA 95687.
Corresponding author: ejfichtner@ucdavis.edu.
320
Proceedings of the Sudden Oak Death Fourth Science Symposium
nicotianae, P. cryptogea, P. gonapodyides, and P. pseudosyringae were baited from
floodwater in madrone pots. Pythium sterilum was also isolated from non-inoculated bay
laurel roots.
The results suggest that the recent mortality at CCSP is caused by P. cinnamomi. It is
unknown when the pathogen was introduced, but a large patch of dead common manzanita
(Arctostaphylos manzanita) and recently killed madrones were noted in the area in 2000. Over
the past decade, P. cinnamomi has also been associated with disease in a natural oak
woodland in southern California (Garbelotto and others 2006); mortality of Ione manzanita
(Arctostaphylos myrtifolia), a rare plant limited to the unusually acidic Ione formation soils
(Swiecki and others 2003); and several areas of madrone, California bay, and manzanita
decline in the San Francisco Bay area.
Additional results of this work have demonstrated that P. cambivora and P. cactorum can
form asymptomatic infections on madrone roots and P. sterilum forms asymptomatic
infections on bay laurel roots. It is not known whether these asymptomatic infections may
develop into root disease under different environmental conditions. Considering that all three
organisms are associated with tree mortality or decline, the hidden transmission of these
organisms as root inhabitants suggests a potential risk to susceptible hosts in both forest and
nursery systems. Additionally, we demonstrated pathogenicity of P. syringae to herbaceous
stems of madrone; however, infectivity by zoospores relied on presence of a pin-prick-sized
wound. To date, P. syringae has not been isolated from madrone in California forest systems;
however, it has been found to infect the related ornamental, Arbutus unedo (strawberry tree),
in nurseries in Spain (Moralejo and others 2008).
The results of this study underscore the risk of pathogen transmission on infested
containerized plants. Pathogens may be transported long distances on non-host plants, either
in infested potting media, or as plant inhabitants associated with asymptomatic infections. For
this study, infested plants were purchased from nurseries specializing in propagation of
natives for habitat restoration plantings. The potential for introducing exotic pathogens
through habitat restoration activities needs to be more widely recognized so that appropriate
phytosanitary procedures can be applied to mitigate this risk to native plant communities.
Literature Cited
Garbelotto, M.; Hüberli, D. and Shaw, D. 2006. First report on an infestation of
Phytophthora cinnamomi in natural oak woodlands of California and its differential impact on
two native oak species. Plant Disease. 90: 685.
Moralejo, E.; Belbahri, L.; Calmin, G.; García-Muñoz, J.A.; Lefort, F. and Descals, E.
2008. Strawberry tree blight in Spain, a new disease caused by various Phytophthora species.
Journal of Phytopathology. 156: 577–587.
Swiecki, T.J.; Bernhardt, E.A. and Garbelotto, M. 2003. First report of root and crown rot
caused by Phytophthora cinnamomi affecting native stands of Arctostaphylos myrtifolia and
A. viscida in California. Plant Disease. 87: 1395.
321
GENERAL TECHNICAL REPORT PSW-GTR-229
The Big Sur Ecological Monitoring Plot
Network: Distribution and Impacts of
Sudden Oak Death in the Santa Lucia
Mountains1
Kerri Frangioso,2 Margaret Metz,2 Allison Wickland,2 David Rizzo,2
and Ross Meentemeyer3
Abstract
The Big Sur area is one of the most ecologically diverse regions in California. Land
preservation efforts are well established in Big Sur, including numerous preserves, state parks,
and the Los Padres National Forest. It appears that no manner of preservation has been able to
protect these wild areas from conservation threats such as exotic species (plants, animals, and
pathogens). Big Sur has provided an exceptional environment to address questions about the
ecological ramifications of Phytophthora ramorum due to the extensiveness of the forests, the
relatively high impact of the disease in this area, and the diversity of environments and
disturbance histories.
High-resolution, digital aerial photography integrated into a GIS was used to map habitat
types and tree mortality associated with P. ramorum in the Big Sur region. This information
was the basis for a model built to randomly generate the location of ecological monitoring
plots. Plots were stratified by forest type (mixed-evergreen and redwood-tanoak), level of tree
mortality, watershed, fire history, and land ownership (public versus private). In 2006 and
2007, we established 280 long-term ecological monitoring plots throughout the region. Within
each 500 m2 circular plot, all stems greater than 1 cm diameter at breast height (dbh) were
identified, measured, mapped, and scrutinized for Phytophthora symptoms and evidence of
other pests. We also quantified the number and identity of regenerating seedlings and
saplings, the percent coverage of each species, and plot-wide canopy height and openness, as
well as topographical descriptors such as elevation, slope, and aspect.
In sum, we collected detailed information on over 13,400 trees throughout the Big Sur region.
Of the 280 plots, 143 are on public land and 137 are on private land. There are 163 mixedevergreen plots and 117 redwood-tanoak plots from Carmel Valley in the north to the
Monterey County line in the south. Eighty of 163 mixed-evergreen plots and 73 of 117
redwood-tanoak plots tested positive for P. ramorum for a total of 153 plots out of the total
280 testing positive for the pathogen responsible for sudden oak death (SOD). Thirty-seven
plots that did not have P. ramorum tested positive for other species of Phytophthora. Twenty
plots had 100 percent infection of tanoak (Lithocarpus densiflorus) and a few plots had 100
percent infection of coast live oak (Quercus agrifolia). In P. ramorum-positive plots, many
host species are living and showing potentially fatal canker or twig symptoms.
The mean standing dead basal area per plot in infested plots compared to uninfested plots is
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
Department of Plant Pathology, University of California, Davis, CA 95616.
3
Department of Geography and Earth Sciences, University of North Carolina at Charlotte, Charlotte,
NC 28223.
Corresponding author: kfrangioso@ucdavis.edu.
322
Proceedings of the Sudden Oak Death Fourth Science Symposium
significantly higher for host species in both forest types. Total dead host basal area in the form
of standing dead stems was 77 percent (p-value = 0.0001) higher in infested redwood-tanoak
plots and 36 percent (p-value = 0.0005) higher in infested mixed-evergreen plots compared to
background mortality levels in uninfested plots. Dead host basal area in the form of downed
coarse woody debris (logs >20 cm diameter) was almost 3 times the amount for tanoak and
1.4 times for host oak species in infested plots compared to background mortality levels in
uninfested plots. This metric not only suggests mortality levels above the standing mortality
measured, but also has implications for nutrient cycling and fire affects.
Mortality from SOD has caused shifts in species abundance; infested stands are increasingly
dominated by species such as California bay laurel (Umbellularia californica) that are not
killed by the disease. There are significantly more bay stems in infested mixed-evergreen
plots than in uninfested mixed-evergreen plots.
Mortality in positive plots was 13 percent for tanoak, 19 percent for coast live oak, and 12
percent for Shreve’s oak (Q. parvula var. shrevei). However, much higher levels of mortality
are occurring in certain stem size classes. For example, in tanoak, over three times the
mortality is occurring in the larger size classes as compared to the smallest stem size class.
These plots provide invaluable information on environment, vegetation, forest structure,
disease level, and site history in areas with and without the disease. Understanding the current
spatial distribution of P. ramorum on the landscape, how this distribution is changing, and the
underlying influences on establishment and spread of P. ramorum will be critical to making
management decisions throughout the state of California. Furthermore, with the arrival of the
2008 fires in Big Sur and our extensive pre-fire dataset throughout our plot monitoring
network, we are ideally situated to learn about the first wildfire in SOD-impacted wildlands.
323
GENERAL TECHNICAL REPORT PSW-GTR-229
Landscape Epidemiology of Species
Diversity Effects on Disease Risk1
S.E. Haas,2 M. Metz,3 K. Frangioso,3 D. Rizzo,3 and R.K.
Meentemeyer2
Abstract
Identifying environmental variables contributing to Phytophthora ramorum spread and
persistence is critical to management and preservation of threatened forest ecosystems. Recent
studies have shown that species diversity can affect disease risk via alterations in transmission
potential between hosts. Most diversity-disease risk studies to date have focused on animals,
with much less attention on generalist plant pathogens in natural ecosystems, and little has
been done to incorporate spatial dimensions of landscape heterogeneity into such studies.
Here, we examine diversity-disease risk in P. ramorum, focusing on the effects of species
diversity within field plots and landscape patterns of disease existence and vegetative
assemblages among plots throughout Big Sur, California. We include ‘force of infection’ in
our regression models to account for inoculum exposure pressure from all known sources of
disease throughout the study area. The analyses revealed a negative relationship between
species richness (number of species; range: 1 to 11) within plots and disease risk. The force of
infection covariate in our model accounts for most of the predictive power of disease risk,
followed by California bay laurel (Umbellularia californica) prevalence per plot. Our findings
agree with other research in multi-host disease systems in that high species diversity is more
likely to decrease than increase disease risk. This result may be occurring in this disease
system through two specific mechanisms—‘encounter reduction’ (reduced encounters
between susceptible and infected host species as additional non-host species are included) and
‘susceptible host regulation’ (a reduction in the number of susceptible hosts as diversity
increases), both of which are the focus of ongoing research. Ongoing research entails using
structural equation modeling (SEM) to further elucidate the complex relationships among
abiotic and biotic landscape heterogeneity variables and disease risk.
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
Center for Applied GIS, UNC Charlotte, Charlotte, NC 28223.
3
Department of Plant Pathology, UC Davis, Davis, CA 95616.
Corresponding author: shaas1@uncc.edu.
324
Proceedings of the Sudden Oak Death Fourth Science Symposium
Effects of Sudden Oak Death on Habitat
Suitability for the California Spotted Owl
(Strix occidentalis occidentalis) in the Big
Sur Ecoregion1
Emily Holland,2 James Hart,3 Kevin Cooper,4 Mark Borchert,4 Kerri
Frangioso,5 David M. Rizzo,5 and Ross K. Meentemeyer2
Abstract
Emerging infectious diseases are increasingly recognized as a major threat to wildlife. The
invasive forest disease sudden oak death (SOD) has recently caused considerable mortality of
oak (Qurecus spp) and tanoak (Lithocarpus densiflorus) trees in the Big Sur ecoregion of
California, which may negatively impact a range oak forest obligate animal species. In this
poster we examine the potential influence of SOD tree mortality on the occurrence of the
declining California spotted owl (Strix occidentalis) in Big Sur forests.
Specifically, we test two alternative hypotheses using a combination of forest field plots, owl
occupancy data, and multi-scale GIS analysis of habitat factors: 1) spotted owls are less likely
to occur in forests with larger amounts of SOD tree mortality due to habitat loss and fewer
food resources for its mammalian prey, or 2) abundant snags and coarse woody debris due to
tree mortality enhances habitat for its prey and thus increases probability of spotted owl
occurrence. Our multi-scale logistic regression modeling indicated that spotted owls are more
likely to occur in forests with greater amounts of tree mortality, after accounting for forest
structure, topography, and fire disturbance variables, and the explanatory power of the tree
mortality-owl presence effect increases with increasing landscape extents up to 400 m radius
measurement areas.
These possibly counterintuitive results suggest that SOD could actually benefit spotted owl
populations, but we hypothesize only in the short term. Over time as this disease-induced
wave of snags and coarse woody debris subsides, the spotted owl may be faced with less
suitable forest habitat and fewer food resources for its prey.
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
Department of Geography and Earth Sciences, University of North Carolina at Charlotte, 9201
University City Blvd, Charlotte, NC 28223.
3
U.S. Fish and Wildlife Service.
4
USDA Forest Service, Los Padres National Forest, 11190 E, Ojai Ave, Ojai, CA 93023.
5
Department of Plant Pathology, University of California, One Shields Avenue, Davis, CA 95616.
Corresponding author: rkmeente@uncc.edu.
325
GENERAL TECHNICAL REPORT PSW-GTR-229
Influence of Nitrogen Fertility on the
Susceptibility of Rhododendrons to
Phytophthora ramorum1
Rita L. Hummel,2 Marianne Elliott,2 Gary Chastagner,2 Robert E.
Riley,2 Kathy Riley,2 and Annie DeBauw2
Introduction
Research information demonstrating the effects of various cultural practices on host
susceptibility to Phytophthora ramorum is generally lacking and thus limits the
development of an integrated approach to managing diseases caused by this pathogen
in irrigated nursery systems. Because rhododendrons and azaleas have accounted for
about 90 percent of the plants associated with P. ramorum-positive nursery finds in
Washington State (as well as being the most important hosts of P. ramorum in
Europe), their management in an irrigated nursery environment is critical to
controlling the spread of this pathogen (Dart and Chastagner 2007). Nitrogen fertility
levels have been reported to increase disease levels in some Phytophthora
pathosystems (Halsall and others 1983), but no data is available for the P. ramorumrhododendron pathosystem.
Methods
During 2008, we investigated the dynamics between nitrogen (N) application rates
and the susceptibility of rhododendron cultivars ‘English Roseum,’ ‘Cunningham’s
White,’ and ‘Compact P.J.M.’ to P. ramorum. Plants were transplanted from 1 gallon
to 3 gallon containers in a medium of 100 percent Douglas-fir (Pseudotsuga
menziesii) bark with Micromax™ incorporated at the rate of 1.75 lbs/yd3, placed on a
gravel nursery bed, and watered as needed with overhead sprinkler irrigation. Before
beginning the experiment, residual fertilizer in the media was depleted and three
treatments of ammonium nitrate fertilizer at 100, 300, and 600 ppm N was applied in
liquid form twice a week to each of eight plants per cultivar starting on June 2. With
each N fertilization, phosphorus in the form of potassium phosphate (100 ppm) and
potassium in the form of potassium sulfate (200 ppm), were applied. Commencing
with fertilizer application, the plants were switched to a drip irrigation system. In
early October, plant growth, visual quality, and leaf color were measured. Color of
the adaxial (upper) leaf surface of two mature leaves from the most recent growth
flush was determined quantitatively with a Minolta CR200b Chroma Meter (Minolta,
Ramsay, New Jersey). The CIELAB coordinates, L*a*b*, were recorded and the
chroma (C*) and hue angle (h) were calculated (McGuire 1992). At the same time,
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
Washington State University, Puyallup Research and Extension Center, 2606 W. Pioneer, Puyallup,
WA 98371.
Corresponding author: hummelrl@wsu.edu.
326
Proceedings of the Sudden Oak Death Fourth Science Symposium
two sets of fully mature, current season leaves from each plant were harvested for 1)
determination of leaf tissue N content and 2) P. ramorum inoculations.
Six detached leaves from each plant were inoculated with zoospores from an NA1
lineage rhododendron isolate of P. ramorum (03-74-N10A-A, from R. x ‘Unique’) by
pipetting a 10 µl drop of suspension with 568,000 zoospores/ml onto the lower leaf
surface. The leaf tissue beneath drops on three leaves was wounded using an insect
pin, while the tissue beneath each drop on the other leaves was left unwounded.
Leaves were incubated in Petri plates with moist filter paper in the dark at 19 to 20 ºC
for 10 days.
Results
As expected, foliage color, shoot growth, plant quality indices, and foliage N levels
increased with N fertility. Observed leaf color correlated with measured leaf color
and plants given higher rates of N were greener than those fertilized at lower rates.
Fertility had no effect on root length or density. Foliar N concentration increased
with N rate. Based on an overall analysis of lesion size after 10 days, there was a
significant difference in the susceptibility of the three cultivars to P. ramorum.
‘Compact P.J.M.’ had the smallest lesions, while ‘English Roseum’ had the largest.
Lesions developed on all the wounded and unwounded inoculation sites on the
‘English Roseum’ and lesion size increased with increasing nitrogen fertility.
Nitrogen fertility had no effect on lesion size on the other two cultivars.
Acknowledgments
This research was supported by the Washington State University Emerging Issues Research
Grant Program.
Literature Cited
Dart, N.L. and Chastagner, G.A. 2007. Estimated economic losses associated with
the destruction of plants due to Phytophthora ramorum quarantine efforts in
Washington State. Online. Plant Health Progress doi:10.1094/PHP-2007-0508-02RS.
Halsall, D.M.; Forrester, R.I. and Moss, T.E. 1983. Effects of nitrogen,
phosphorus and calcium nutrition on growth of eucalypt seedlings and on the
expression of disease associated with Phytophthora cinnamomi infection. Australian
Journal of Botany. 31: 341–355.
McGuire, R.G. 1992. Reporting of objective color measurements. Horticultural
Science. 27: 1254–1255.
327
GENERAL TECHNICAL REPORT PSW-GTR-229
Stream Baiting for Sudden Oak Death:
Fluvial Transport and Ecohydrology of the
Invasive Plant Pathogen Phytophthora
ramorum in Western Washington State1
Regina Johnson2
Abstract
Phytophthora ramorum, a member of the water molds (Oomycota), spreads in water and
survives unfavorable conditions in soil. The pathogen has been shown to travel 15 m in
windblown rain, and as much as 7 km in flowing water, and to survive up to 8 months in soil.
In forest settings in California, windblown rain poses a major dispersal agent for P. ramorum,
picking up spores from tree tops. In Washington State, P. ramorum is a pathogen in nursery
settings rather than forests. Nursery stock on which spores are produced tend to be very small
plants, usually less than 1 m tall, and spores tend to be dropped to the ground rather than
picked up by wind and rain. This dispersal pattern, along with the ability to survive in soil and
to be transported in water, suggests that in nursery settings, soil, soil water, and streams may
be critical dispersal agents for this invasive pathogen.
Phytophthora ramorum has been found on plants and in soil, potting mix, and surface waters
in nurseries in Washington State. Two years of detection efforts by the Washington State
Department of Agriculture (WSDA) are reviewed in this study. The WSDA baited six streams
in 2006, and eight in 2007, for a total of 11 different streams. Of these streams, P. ramorum
has been found in three. Positive stream baits in all three streams show a correlation with
rising temperatures and decreasing precipitation in spring, particularly in April.
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
The Evergreen State College, 2700 Evergreen Parkway NW, Olympia, WA 98505.
Corresponding author: reg@madronas.net.
328
Proceedings of the Sudden Oak Death Fourth Science Symposium
Precip. mm
Temp. C
20
Harst. Positives
18
Samm. Positives
Precipitation, mm
500
Tukwila Positive
16
14
400
12
10
300
8
200
6
4
100
Temperature C, # Positives
600
2
1
1 2
-0
8
9
10
11
8
6
7
5
4
3
11
1
1 2
-0
7
2
8
9
10
7
6
5
4
3
0
1
1 2
-0
6
2
10
-0
5
11
0
Figure 1—Precipitation, temperature, and positive stream baits for R Stream
(Harstine), the Sammamish River, and Tukwila soil drainage ditch, 2006-07. Climate
data for R Stream area from Western Regional Climate Center,
http://www.wrcc.dri.edu. Climate data for Sammamish River/Tukwila ditch area are
similar in pattern.
Jan
6%
Feb
6%
Mar
6%
May
24%
Apr
58%
Figure 2— Positives in R Stream by month, 2006-2007.
Five P. ramorum-positive nurseries are compared by soil and hydrologic factors. Two are
associated with positive streams, while the other three failed to produce P. ramorum-positive
stream baits. Of these five nurseries, only one is associated with multiple positive stream baits
over the 2-year period. This nursery has soil and hydrologic features distinct from all other
nurseries studied. These features may be diagnostic for sites conducive to the escape of P.
ramorum from the nursery environment to establishment in the wider landscape, in particular,
329
GENERAL TECHNICAL REPORT PSW-GTR-229
coarse, gravelly, shallow, sloped soil, and a low-order, high-gradient, gravel-bed stream.
Deep, fine, biologically active soils filter microbes, while shallow, coarse, biologically
inactive soils allow microbes to pass through (Brady and Weil 2002). Shallow, coarse,
Harstine soil may allow propagules of P. ramorum to pass through in soil water from the
nursery into the stream. All other soils in this study are deep, fine-textured soils, which would
be expected to filter and trap microbes in soil water.
Hardpans are known to concentrate and transport spores of P. cinnamomi, potentially as far as
120 m (Shea and others 1983, Kinal and others 1993). Both R Stream and the Tukwila soil
drainage ditch positives are associated with a shallow hardpan. A natural hardpan underlies
Harstine soil, while the Tukwila soil nursery has created an artificial hardpan by building
raised display beds of composted shavings on top of muck soil. One soil-positive display bed
is sloped towards the stream-positive drainage ditch. Shallow hardpan, coarse soil, and slope
appear to be conducive to escape of P. ramorum from nurseries into streams. More data are
necessary to examine the relative effects of these three factors.
Table 1—Nursery soil types with selected characteristics. Soil data from
NRCS’s Web Soil Survey.
Soil Series Name
General type
Slope, %
Hardpan
AWC* Hydrologic Coarse
group** fractions
Harstine (R Stream), repeat gravelly sandy loam,
stream positive
glacial till
very fine sandy loam/silt
Newberg/Nooksack
loam, alluvial
6
yes
0
no
artificial, at
soil/shavings
interface
artificial, at
soil/shavings
interface
no
low
very
high/high
C
15-25%
B/C
0
Terric Medisaprist (under
composted shavings)
muck
0
moderate
ponded
Tukwila (under composted
shavings), single stream
positive, Samm. tributary
muck
0
high
D (ponded)
Woodinville
silty clay loam, alluvial
0
high
D (flooded)
* Available water content.
** Categorizes soil infiltration rate and runoff potential, with A = high infiltration/low runoff
and D = low infiltration/high runoff.
While the nursery itself has apparently been cleared of P. ramorum, and no infested riparian
plant material has been found, the stream continues to produce positive stream baits. Survival
in damp soil has been shown to be sufficiently long for P. ramorum to survive Washington’s
dry season in stream sediments (Fitchner and others 2007). P. ramorum in the form of
decomposing infected plant material could be stored in sediments in the channel, along the
banks, and in floodplains, and could still be infective when remobilized by the stream. Stream
sediments could be acting as reservoirs of infective material in the absence of infected plant
hosts.
Acknowledgments
The author thanks Evergreen State College, Masters in Environmental Studies Program; Paul
Butler, Ph.D. thesis advisor and Paul Przybylowicz, Ph.D. reader; and Washington State
Department of Agriculture, Jennifer Falacy, thesis technical advisor.
330
0
0
0
Proceedings of the Sudden Oak Death Fourth Science Symposium
Literature Cited
Brady, N.C. and Weil, R.R. 2002. The nature and properties of soils. 13th edition. Upper
Saddle River, NJ.: Prentice Hall. 960 p.
Fichtner, E.J.; Lynch, S.C. and Rizzo, D.M. 2007. Detection, distribution, survival, and
sporulation of Phytophthora ramorum in a California redwood-tanoak forest soil.
Phytopathology. 97: 1366–1375.
Kinal, J.; Shearer, B.L. and Fairman, R.G. 1993. Dispersal of Phytophthora cinnamomi
through lateritic soil by laterally flowing subsurface water. Plant Disease. 77: 1085–1090.
Shea, S.R.; Shearer, B.L.; Tippett, J. and Deegan, P.M. 1983. Distribution, reproduction,
and movement of Phytophthora cinnamomi on sites highly conducive to jarrah dieback in
South Western Australia. Plant Disease. 67: 970–973.
331
GENERAL TECHNICAL REPORT PSW-GTR-229
Evaluation of Diurnal Rhythms in
Phytophthora1
Takao Kasuga2 and Mai Bui2
Abstract
A daily rhythmic activity cycle, or circadian rhythm, is an endogenously generated 24-hour
periodicity, and can be entrained by external cues, such as daylight. These rhythms allow
organisms to anticipate and physiologically prepare for precise and regular environmental
changes. Many organisms, including plants, animals, fungi, and cyanobacteria, are known to
display circadian cycles. In the fungal kingdom, rhythms in spore development and discharge
are widespread, which indicate a selective advantage for regulation of these events at specific
times of the day.
It is not known if oomycete pathogens possess endogenous circadian rhythm or whether they
utilize metabolic cues from host plants to generate diurnal periodicity. It has, however, been
shown that in the Phytophthora capsici-pepper hydroponic system, sporangium production
and zoospore release were cyclic; sporangium production reaches a peak at 4 p.m., whereas
zoospore release takes place at a maximum rate 2 hours after dark (Nielson and others 2006).
Diurnal periodicity has also been reported for sporulation of downy mildew pathogens
Plasmopara viticola (Rumbolz and others 2002, Yarwood 1937) and Bremia Lactucae
(Nordskog and others 2007).
Furthermore, it is known that, for example, in fungi as much as 20 percent of the total gene
transcripts display circadian periodicity. Because of this confounding effect, disregarding
circadian periodicity while conducting global mRNA profiling or proteomics research, can
potentially lead to misinterpretation of high throughput datasets. This has urged us to evaluate
diurnal rhythmicity of P. ramorum and its potential link to pathogenicity using microarray
mRNA profiling.
We grew P. ramorum under a diurnal photo cycle (12 hours light : 12 hours dark) for 6 days,
then shifted to a 24 hour constant dark condition. Persistence of 24-hour periodicity in mRNA
expression was then evaluated using fast Fourier transformation analysis. Out of 15,495 gene
expression profiles obtained by P. ramorum NimbleGen microarray, only 94 genes (0.6
percent) showed persistent 24 hour diurnal rhythmicity. Because genes showing circadian
rhythmicity in P. ramorum were much less than those found in ascomycete fungi, it is not
clear if Phytophthora possesses an endogenous circadian system. Given the assumption that
diurnal rhythmicity is essential to survival, P. ramorum might use cues of diurnal cycle from
host plants. Phytophthora might be a circadian parasite.
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
Crop Pathology & Genetics Research Unit, USDA ARS, 382 Hutchison Hall, University of California,
Davis, CA 95616.
Corresponding author: tkasuga@ucdavis.edu.
332
Proceedings of the Sudden Oak Death Fourth Science Symposium
Literature Cited
Nielsen, C.J.; Ferrin, D.M. and Stanghellini, M.E. 2006. Cyclic production of sporangia
and zoospores by Phytophthora capsici on pepper root in hydroponic culture. Canadian
Journal of Plant Pathology. 28: 461–466.
Nordskog, B.; Gadoury, D.M.; Seem, R.C. and Hermansen, A. 2007. Impact of diurnal
periodicity, temperature, and light on sporulation of Bremia lactucae. Phytopathology. 97:
979–986.
Rumbolz, J.; Wirtz, S.; Kassemeyer, H.H.; Guggenheim, R.; Schäfer, E. and Büche, C.
2002. Sporulation of Plasmopara viticola: differentiation and light regulation. Plant Biology.
4: 413–422.
Yarwood, C.E. 1937. The relation of light to the diurnal cycle of sporulation of certain
downy mildews. Journal of Agricultural Research. 54: 365–373.
333
GENERAL TECHNICAL REPORT PSW-GTR-229
OakMapper 2.0: Distributed Participatory
Sensing for Monitoring and Management of
Sudden Oak Death1
Maggi Kelly,2 John Connors,2 and Shufei Lei2
Background
Public interest in sudden oak death (SOD), caused by Phytophthora ramorum,
remains high as the disease continues to spread and impact more areas. Early in the
infestation, information from active members of the public was key in locating new
areas of infestation across the state. The California Oak Mortality Task Force
(COMTF), arborists, and university researchers were repeatedly contacted with
reports of new areas of suspected infestations. In response to this concern from the
public, we created a website in 2001 where visitors could submit the locations of
trees that were potentially infected. This site, OakMapper (www.oakmapper.org), has
had thousands of visitors who have submitted hundreds of point locations of trees
suspected of having the disease. In addition to this functionality, over time the first
version of the OakMapper served as a clearinghouse for four SOD-related, spatial
resources: 1) Google Maps, 2) Google Earth, 3) ESRI ArcIMS, and 4) static maps.
All of these resources were dependent upon a project administrator to manually
update their source data and reload the content to the website on a quarterly basis.
The OakMapper webGIS application is our comprehensive database and cartographic
portal, containing all SOD data available for public viewing. In October 2008, we
launched the second version of our webGIS, OakMapper 2.0, offering a more
dynamic, customizable, and user-driven cartographic environment that is built on a
combination of open-source and proprietary software (Kearns and others 2003,
USDOC 2002). OakMapper 2.0 allows user-specific interactions – including scaledependent zooming, customized map creation, hyperlinked photography, and
querying functions – using the spatial database PostGIS. The webGIS site also allows
users to report trees that might have the disease so that follow-up sampling can take
place. The development of web-based efforts continues to prove effective in
communicating SOD information to researchers, regulators, and the general public by
providing a readily available avenue for viewing, searching, querying, and exporting
data and maps. The ultimate goal of the OakMapper webGIS is to empower
stakeholders to participate in disease monitoring. To this end, the application is
designed with non-GIS experts in mind. An online form is used to gather reports of
potential SOD sightings by allowing users to: 1) select a host and visible SOD
symptoms (chosen from pictures and explanations that aid in identification), 2) enter
information about their professional background, and 3) submit the location of the
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
Department of Environmental Sciences, Policy and Management, University of California, Berkeley.
145 Mulford Hall Number 3114, Berkeley, CA 94720.
Corresponding author. maggi@berkeley.edu.
334
Proceedings of the Sudden Oak Death Fourth Science Symposium
tree (GPS coordinates, addresses, or location on map). The numerous submissions to
date have demonstrated the success of citizen-generated data in widening the
sampling effort for this disease (Kelly and Tuxen 2003, Kelly and others 2004).
The New OakMapper
OakMapper 2.0 integrates the features of OakMapper 1.0 into one package and then
further extends to other new features. As new open-source tools became available,
we were interested in migrating to these more flexible solutions. The migration
process begins with consolidating disparate data storage formats and sources, such as
shapefile, MS Access, and Excel, into a single format and data source for which we
selected SQL in the open-source PostgresSQL. The existing data is then migrated
into a spatial database, enabled by PostGIS. PostGIS, which is an open-source spatial
database, allows us to perform spatial data query and analysis. (Currently,
OakMapper 2.0 is not utilizing the full features of PostGIS; this is set for future
development.)
OakMapper 1.0 had four distinct and primary components: static maps, ESRI
ArcIMS, Google Maps API, and Google Earth KML/KMZ. Instead of having them in
an integrated system, the front page of OakMapper 1.0 functions like a portal web
page for each of these four components. As a result, there is no navigation and
interaction between these four components within the OakMapper 1.0 site.
OakMapper 2.0 first integrates these four components by providing a navigation
menu at the top. The navigation system allows users to travel back and forth among
these components easily and provides a consistent feel and experience throughout the
site. OakMapper 2.0 also allows different components to interact with one another.
For example, the static maps can be selected for download using the Google Maps
API download tool. Also, when you submit a point to the system via the Google
Maps API, the Google Earth KML data file will be automatically updated.
OakMapper 2.0 (fig. 1) allows
any users to come to the system
and submit new findings of
SOD to the database. Designing
the system for the general
public, the migration follows the
user-centered design philosophy
to achieve ease of use for end
users. When reporting a
suspected case of SOD, users
simply 1) draw a point or a
polygon on the Google Map and
2) enter relevant information,
such as descriptions and even
pictures about the new SOD
find. This easy-to-use system is
built to encourage community
Figure 1—OakMapper 2.0 website front page.
participation in recording more
SOD occurrences, so that spread
can be tracked more efficiently. And given that users' submissions are open to the
general public, the public can be alerted about the new occurrences of SOD. The
335
GENERAL TECHNICAL REPORT PSW-GTR-229
most recent SOD submissions will be displayed on the homepage, so that users can
view the most recent activity on the site. The interaction between these features is
enabled by their shared database.
OakMapper 2.0 allows users to register into the system so that they can keep track of
their SOD submissions. Given that users might want to modify the descriptions or
other information of their SOD submissions, registered users are provided with tools
to edit their submissions. Registered users can also provide comments on SOD
submissions. The commenting features of OakMapper 2.0 will facilitate more
information generation and community building. Users can comment on the severity
of SOD submissions. Like the submissions of SOD, users can keep track of and edit
their submitted comments in the “My Account” section.
To improve the system’s responsiveness to users’ activities on the OakMapper 2.0
site, the system sends a confirmation email to the users when they register and when
they submit an observation of SOD. The confirmation email will also contain the
most recent SOD submissions and the most recent comments, which link back to
OakMapper 2.0 for further exploration. RSS feed is a familiar tool in the Web 2.0
world. The GeoRSS standard provides a way to integrate RSS feeds with location
information. OakMapper 2.0 generates GeoRSS feeds so that feed readers with
spatial awareness can take advantage of the RSS feed of SOD submissions. The
standard GeoRSS format allows the SOD data to be integrated with other web-map
mashup applications.
OakMapper serves as an important resource for researchers to access the most up-todate maps of confirmed cases of SOD. Our new WebGIS, built on ESRI ArcGIS
Server, utilizes ArcSDE to reference the PostGIS spatial database to display the most
up-to-date data available. This new structure ensures that users have access to all
confirmed points and frees the site administrator from manually creating dozens of
static maps.
Sign Up on OakMapper
The official map of sudden oak death in California shows only a few hundred
individual trees with the disease. This is because of the time and expense required to
officially confirm the presence of P. ramorum; the California Department of Food
and Agriculture and the University of California perform this confirmation process
on samples collected statewide. This map of individual trees does not show the
complete extent of oak mortality statewide, and we are interested in getting public
help in mapping other pockets of oak mortality that are not shown on the official
map. Not all of these areas can or will be officially confirmed to have the disease, but
we are interested in further defining where oak mortality exists, with your help. For
example, there are many clusters of oak mortality in the East Bay Regional Parks that
have not yet been mapped. OakMapper 2.0 can help. We would like you to use this
tool to map areas where you see pockets of oak mortality that might be connected to
SOD. We hope that this model of data acquisition, storage, analysis, and
dissemination will be more widely used in forest health management in particular,
and in natural resource management in general, while proponents of such a system
will remain cognizant of the potential challenges.
336
Proceedings of the Sudden Oak Death Fourth Science Symposium
Literature Cited
Kearns, F.R.; Kelly, M. and Tuxen, K.A. 2003. Everything happens somewhere: using
webGIS as a tool for sustainable natural resource management. Frontiers in Ecology and the
Environment. 1(10): 541–548.
Kelly, M.; Tuxen, K.A. and Kearns, F.R. 2004. Geospatial informatics for management of a
new forest disease: Sudden Oak Death. Photogrammetric Engineering and Remote Sensing.
70(9): 1001–1004.
Kelly, N.M. and Tuxen, K. 2003. WebGIS for monitoring “sudden oak death” in coastal
California. Computers, Environment and Urban Systems. 27(5): 527–547.
USDOC [U.S. Department of Commerce]. 2002. A nation online: how Americans are
expanding their use of the Internet. United States Department of Commerce, Economics and
Statistics Administration, and National Telecommunications and Information Administration.
February 2002. http://www.ntia.doc.gov/ntiahome/dn/html/toc.htm. [August 08, 2009].
337
GENERAL TECHNICAL REPORT PSW-GTR-229
Quantifying Large-Scale Impacts of Sudden
Oak Death on Carbon Loss in the Big Sur
Basin Complex Fire: Upscaling From the
Plot to Region1
Sanjay Lamsal,2 Ross K. Meentemeyer,2 Qingmin Meng,2 Margaret
Metz,3 Richard Cobb,3 Kerri Frangioso,3 and David M. Rizzo3
Abstract
In California forests, sudden oak death- (SOD) related mortality and frequent wildfires have
emerged as an important forest management concern. The forests of Big Sur are among the
most impacted by SOD. Widespread tree mortality reduces biome production and carbon
uptake, and increases future carbon emissions from decay and burning of coarse woody debris
and dead trees. We hypothesize that SOD has a positive feedback on fire intensity and
increases future carbon emissions, but the effects may be confounded by site physiography,
fire characteristics (direction, intensity), and weather conditions. Our goal is to assess the
interaction between SOD mortality, forest fire characteristics, and SOD contribution to
aboveground biomass/carbon losses in the Big Sur region. We surveyed 280 plots, and
estimated volume of coarse woody debris and biomass/carbon stored within live trees using
their diameter at breast height. Post fire effects were surveyed in 61 burned plots to assess fire
characteristics and their effects on coarse woody debris. Analyses are ongoing to estimate
carbon losses from coarse woody debris across plots with different levels of mortality. Future
surveys will focus on estimating SOD contributions to biomass/carbon losses from the plots
and quantifying the large scale effects across the Big Sur region. Ignoring the effects of SOD
on carbon dynamics and failure to account for the losses may overestimate the potential for
forests to offset anthropogenic CO2.
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
Department of Geography and Earth Sciences, University of North Carolina at Charlotte, Charlotte,
NC 28223.
3
Department of Plant Pathology, University of California at Davis, Davis, CA 95616.
338
Proceedings of the Sudden Oak Death Fourth Science Symposium
The California Oak Mortality Task Force and
Phytophthora ramorum Outreach1
Chris Lee,2 Janice Alexander,3 and Katie Palmieri4
Abstract
Since 2000, the California Oak Mortality Task Force (COMTF) has coordinated a
comprehensive program of research, management, monitoring, education, and public policy
related to Phytophthora ramorum and its impacts. In support of the COMTF mission, all five
of these disciplines have education and outreach components. This education and outreach
represents one of the first and most wide-reaching efforts to coordinate and provide
information about a forest disease for disparate audiences, a task made even more complex by
P. ramorum’s occurrence in nurseries as well as forests.
The COMTF accomplishes these goals with the help of two outreach coordinators, a public
information officer, an outreach associate, and a webmaster. These personnel are assisted on
specific projects by additional university and agency staff around the state. Their efforts
provide up-to-date science-based P. ramorum-related information to the 14 infested counties
in California, as well as throughout non-infested areas in the state and the country, using
educational materials via every major medium. This information is directed to the general
public, land managers, other natural resource professionals, affected industries, regulators,
policy makers, news media, educators, and scientific researchers. Additionally, the outreach
and education staff assists in linking P. ramorum researchers throughout the world to each
other by facilitating meetings, scientific symposia, and teleconferences.
One successful example of COMTF outreach and education work involves treatment and
training workshops provided every year in a variety of communities throughout California.
Some of these workshops, given in cooperation with the University of California, Berkeley
Garbelotto laboratory, give participants an opportunity to closely observe techniques for
treating trees and landscapes to prevent or lessen the impacts of P. ramorum on their
properties or in their communities. Another series of repeating workshops brings together P.
ramorum researchers, educators, and ecologists from throughout California to give varied
audiences the latest information on P. ramorum biology, distribution, monitoring,
management, and regulations.
Other notable examples of outreach and education efforts include the creation and
maintenance of a website that serves as the main nexus for easy-to-find information on P.
ramorum; the facilitation of numerous local meetings to discuss P. ramorum impacts in
specific communities; a monthly newsletter that summarizes the latest P. ramorum-related
findings and information; the compilation of P. ramorum information and the coordination of
monitoring efforts with local tribes; communication with media; and the fielding of daily
phone calls concerning P. ramorum and related forest health issues from the public. Periodic
self-assessment through public surveys helps to refine outreach efforts and guide the
development of future efforts so that the dissemination of new information improves in
tandem with our own increased knowledge of the pathogen.
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
University of California Cooperative Extension, 5630 South Broadway, Eureka, CA 95503.
3
UC Cooperative Extension, Marin County, 1682 Novato Blvd, Novato, CA 94947.
4
UC Berkeley, Center for Forestry, Mulford Hall, Berkeley, CA 94720.
Corresponding author: cale@ucdavis.edu.
339
GENERAL TECHNICAL REPORT PSW-GTR-229
Oregon’s Grower Assisted Inspection
Program: an Audit-Based System to
Manage Phytophthora Diseases1
Melissa Lujan,2 Gary McAninch,2 and Nancy Osterbauer2
Introduction
Invasive plant pathogens are a tremendous threat to forest health. Phytophthora
ramorum, an exotic fungus-like organism, attacks 118 plant species and kills mature
oak, tanoak, and beech trees. It is established in 14 counties in California and is
found in a limited area near Brookings, Oregon. The pathogen is also a problem in
nurseries, where it can infect plants and infest soil and irrigation water. In Europe, it
has been shown that P. ramorum can spread from infected nursery stock into natural
landscapes. A federal quarantine regulates the movement of P. ramorum-susceptible
plants within the U.S. However, the pathogen continues to be detected in plants
moving through the nursery trade. The Oregon Department of Agriculture (ODA)
worked with Oregon’s nursery industry and others to develop a systematic approach
to managing Phytophthora problems, including P. ramorum, in nurseries. The
voluntary Grower Assisted Inspection Program (GAIP) is designed to compliment
the federal quarantine by enlisting the cooperation of nurseries in preventing the
spread of P. ramorum through the movement of infected plants. The nurseries do so
by adopting best practices for Phytophthora disease management.
Program Structure
Education Requirements
With the help of Oregon State University (OSU) and the U.S. Department of
Agriculture, Agricultural Research Service (USDA ARS), a bilingual online training
course (http://ecampus.oregonstate.edu/phytophthora) that describes Phytophthora
biology, best management practices for Phytophthoras in nurseries, and P. ramorum
specifically was developed. All GAIP participants had to take and pass this course. A
workshop was held to train nursery workers to identify Phytophthora-infected plants
in the field and test them using the Phytophthora Alert®-LF field kit (Neogen Europe,
Ltd.).
Hazard Analysis
Each nursery was required to review their production and procurement processes to
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
Oregon Department of Agriculture, 635 Capitol St NE, Salem, OR 97301.
Corresponding author: nosterba@oda.state.or.us.
340
Proceedings of the Sudden Oak Death Fourth Science Symposium
determine where Phytophthora could be introduced into their operation. OSU and
ARS research identified three critical control points (CCP) where a Phytophthora
could be introduced into a nursery - water, soil and/or potting media, and used
containers (Parke and others 2008). The ODA identified a fourth CCP for P.
ramorum - incoming plants. Nurseries were encouraged to include other CCP unique
to their operation.
Mitigation Manuals
Once all CCP were identified, each nursery had to identify best management
practices that would mitigate the risk of Phytophthora being introduced at that CCP.
Each practice had to be auditable through documentation, interviews, or observations.
The nursery documented these practices in a mitigation manual that was then
reviewed by the ODA.
Nursery Audits
Once their mitigation manual was accepted, each nursery was audited at least three
times per year. The first audit ensured the nursery had the necessary infrastructure to
carry out the best practices outlined in their manual. The second and third audits were
done to verify the nursery was actually following their manual. Host and associated
plants (HAP) were inspected and sampled during the second and third audits.
Corrective Actions
If a nursery is out of compliance with their mitigation manual, they must implement a
corrective action or be suspended from the GAIP. Major non-compliance issues
result in immediate suspension, while minor non-compliance issues must be
corrected within a set time period or the nursery will be suspended. Once additional
audits verify corrective action has been taken, the suspension is lifted.
Quantifying the Impact of GAIP
Eighteen nurseries volunteered to participate in the GAIP. Although all nurseries are
welcome to participate, education and outreach efforts focused primarily on growers
of Rhododendron and Camellia, the hosts reportedly infected with P. ramorum the
most often (Tubajika and others 2006).
To quantify the impact of the GAIP, we used the results of the federal certification
survey (7 CFR 301.92), which requires all growers shipping HAP interstate to be
inspected and have at least 40 samples tested annually for P. ramorum. The ODA
surveys Rhododendron and Camellia growers twice each year as a matter of internal
policy.
During each inspection, samples were collected if suspicious symptoms were
observed. One sample consisted of five individual, symptomatic leaves collected
from a single plant. Sampled plants were flagged for later identification. Samples
were initially screened for Phytophthora using a commercially available ELISA kit
(Agdia, Inc., Elkhart, IN). Samples positive with ELISA were then tested using the
341
GENERAL TECHNICAL REPORT PSW-GTR-229
USDA-approved molecular diagnostic protocols
(http://www.Aphis.usda.gov/plant_health/plant_pest_info/pram/protocols.shtml).
Results and Discussion
In 2007, the ODA inspected all growers/shippers of P. ramorum-susceptible plants to
establish a baseline for aerial Phytophthora species in Oregon nurseries.
Phytophthora were detected in 42.0 percent of the 754 sites surveyed, with P.
ramorum detected at three sites. Of 29,665 plant samples tested, 1,480 were infected
with a Phytophthora and four of those with P. ramorum. This showed that aerial
Phytophthora are a common disease problem in Oregon nurseries, while P. ramorum
is present at a very low level.
In 2007, Phytophthora was detected at 17 of the 18 GAIP nurseries (94.4 percent).
The percentage of samples found infected with Phytophthora was used as a measure
of the level of Phytophthora disease present in each nursery. In 2007, the average
level of disease detected within the nurseries was 14.6 percent. In 2008,
Phytophthora was detected at 16 of the 18 GAIP nurseries (88.9 percent), while the
average level of disease detected was 14.9 percent. From 2007 to 2008, the amount of
disease detected decreased significantly at three nurseries, remained the same at 11
nurseries, and increased significantly at four nurseries (fig. 1, p = 0.10). As of June 2,
2009, Phytophthora was detected in five of seven nurseries surveyed (71.4 percent),
while the average level of disease detected was 8.0 percent. From 2008 to 2009, the
amount of disease detected decreased significantly at four of the seven nurseries
surveyed. In the remaining nurseries, the disease level remained the same. Many of
the volunteers were adopting new best practices at their sites in 2008. We believe the
2009 survey results will provide a better picture of the impact of the GAIP on aerial
Phytophthora species within these nurseries.
We also examined which hosts were found infected within the GAIP nurseries. In
2007, Phytophthora was detected on 17 host genera. Phytophthora was found on
Rhododendron and Pieris the most often, representing 74 percent and 7 percent of all
infected samples, respectively. In 2008, Phytophthora was detected on fewer host
genera (13). Again, Phytophthora was detected on Rhododendron (82 percent of all
infected samples) and Pieris (7 percent of all infected samples) the most often.
Preliminary results from 2009 support these earlier findings. Tubajika and others
(2006) reported Rhododendron and Camellia were at highest risk for spreading P.
ramorum. Our results suggest that in Oregon, disease management and outreach
efforts would be better focused on Rhododendron and Pieris growers.
Further research by OSU and ARS scientists has shown that there is a substantial
endemic Phytophthora population present in many nurseries’ soil substrate (Jennifer
Parke, Oregon State University, Corvallis, OR, personal communication). Many of
the practices adopted by the GAIP nurseries, such as preventing contamination of
potting media with native soil, address this issue indirectly. Further work needs to be
done to identify best practices that directly address the issue of endemic
Phytophthora populations in soil. Therefore, it may take several growing seasons for
new, best practices to have an impact at nurseries with significant, endemic
populations of Phytophthora in their soil substrate.
342
Proceedings of the Sudden Oak Death Fourth Science Symposium
50
45
% Infected Samples
40
35
2007
2008
2009
30
25
20
15
10
5
xx5136
xx0753
xx6010
xx1386
xx4528
xx8738
xx0672
xx2739
xx8610
xx9975
xx9488
xx9749-s
xx9749-g
xx9749-d
xx4677
xx3432
xx3484
xx5065
0
License #
Figure 1—The percentage of aerial Phytophthora disease present at GAIP nurseries in 2007,
2008, and 2009.
Literature Cited
Parke, J.L.; Grunwald, N.; Lewis, C. and Fieland, V. 2008. A systems approach for
managing Phytophthora diseases in production nurseries. Phytopathology. 98: S121.
Tubajika, K.M.; Bulluck, R.; Shiel, P.J.; Scott, S.E. and Sawyer, A.J. 2006. The
occurrence of Phytophthora ramorum in nursery stock in California, Oregon, and Washington
states. Online. Plant Health Progress doi:10.1094/PHP-2006-0315-02-RS.
343
GENERAL TECHNICAL REPORT PSW-GTR-229
Effect of Plant Sterols and Tannins on
Phytophthora ramorum Growth and
Sporulation1
Daniel Manter,2 Eli Kolodny,2 Rick Kelsey,3 and Pilar GonzálezHernández4
Abstract
The acquisition of plant sterols, mediated via elicitins, is required for growth and sporulation
of Phytophthora spp. In this study, we examined the effect of plant sterols and tannins on
growth and sporulation of Phytophthora ramorum. When ground leaf tissue was added to
growth media, P. ramorum growth and sporulation was greatest on California bay laurel
(Umbellularia californica) as compared to either California black oak (Quercus kelloggii) or
Oregon white oak (Q. garryana), which is in agreement with field observations. However,
when purified foliar sterol extracts were added to the media, no difference in growth and
sporulation of P. ramorum was observed, suggesting the presence of an inhibitor in the
foliage of the two oak species. Tannins are polyphenolic compounds that have the ability to
precipitate proteins and are found within a wide array of plants, particularly oak species.
Foliar tannins from all three plant species were able to bind and precipitate elicitins, and a
linear relationship was observed between the amount of elictin removed (precipitated) by the
tannins and P. ramorum growth and sporulation. Based on these studies, we suggest that the
higher tannin content of oaks inhibits P. ramorum growth and sporulation by inactivating
elicitins and sterol acquisition by P. ramorum.
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
USDA ARS, Plant Nutrient Research Unit, 2150 Centre Ave., Building D, Suite 100, Fort Collins, CO
80526.
3
USDA Forest Service, PNW Research Station, Corvallis, OR 97331.
4
Santiago de Compostela University, Lugo, Spain.
Corresponding author: daniel.manter@ars.usda.gov.
344
Proceedings of the Sudden Oak Death Fourth Science Symposium
The Santa Lucia Preserve: A Case Study of
Community-Based Sudden Oak Death
Sampling Blitzes1
Cheryl M. McCormick2 and Matteo Garbelotto3
Abstract
The Santa Lucia Preserve is a 20,000 acre, private, upscale conservation and limited
development project in Carmel, California. In early 2007, Preserve residents, in collaboration
with Dr. Matteo Garbelotto, developed the concept of a “SOD (sudden oak death) Blitz,”
whereby communities conduct a day-long sampling of California bay laurel (Umbellularia
californica) leaves, which are subsequently analyzed for the presence of Phytophthora
ramorum by Dr. Garbelotto’s plant pathology laboratory at the University of California (UC),
Berkeley. California’s first SOD Blitz (coined “Bay Watch”) was held on the Preserve in May
2008 and was attended by 67 Preserve landowners, certified arborists, and Preserve staff.
Outcomes of the event included a map of the distribution of SOD on the Preserve, a
comprehensive SOD management plan for the property, a series of SOD workshops, a “SOD
Blitz Manual,” and a prioritized strategy for managing SOD on the Preserve.
By partnering with Preserve landowners, the Santa Lucia Conservancy has developed a highly
efficient protocol for successfully organizing and implementing a large-scale SOD Blitz. A
team of landowner “Blitz Captains” is responsible for educating their neighbors about SOD
and the Conservancy’s management efforts, encouraging participation in Blitz events, and
participation in numerous pre-event activities, such as sample protocol training, pre-flagging
California bay laurel trees for sampling, and assembling sample kits. This ‘peer-to-peer’
education and outreach is highly effective in motivating landowners to participate in SOD
Blitz events, and also has the added benefit of revealing critical information gaps that exist
within the community, as Preserve community members are more likely to share their
concerns and needs with each other rather than with a member of the Conservancy staff. This
information, in turn, assists the Conservancy in developing workshops, educational resources,
and priorities for management. A total of 253 samples were collected during the inaugural
2008 SOD Blitz, of which 42 (18.2 percent) tested positive for P. ramorum using
conventional microbial and molecular techniques. Because each sample was referenced by a
spatial coordinate transferred from a marked paper map to a GIS database using ArcView 9.3
(ESRI®), positive results can be mapped as discrete points buffered by a zone representing the
average dispersal distance of P. ramorum. The area within these buffered distances represents
the area within which active management should be considered.
As a direct result of the SOD Blitz efforts, an early detection and rapid response (ED&RR)
team effectively eradicated two nascent foci of P. ramorum infection in previously uninfested
habitats on the eastern portion of the Preserve. Landowners whose property was determined to
contain a positive result were advised to consider a two-pronged approach to managing SOD,
involving preventative treatment of bark host species (namely, coast live oaks, Quercus
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
Santa Lucia Conservancy, 26700 Rancho San Carlos Road, Carmel, CA 93923.
3
Department of Environmental Science, Policy, and Management, University of California, Berkeley,
CA 84720.
Corresponding author: cm.conservation@gmail.com.
345
GENERAL TECHNICAL REPORT PSW-GTR-229
agrifolia) with Agri-Fos/Pentrabark® and selective removal and/or pruning of California bay
laurel trees.
The Preserve’s second annual event (“SOD-ZILLA 2009”) was held on March 21, 2009, and
was attended by 34 members of the Preserve community. Results are expected from the
Garbelotto plant pathology laboratory at UC Berkeley in September 2009. A map depicting
changes in the distribution of SOD on the Preserve from 2008 to 2009 will be produced and
incorporated into the Conservancy’s SOD management plan.
The Santa Lucia Conservancy’s successful SOD Blitz protocols and organizational structure
served as a template for a region-wide SOD Blitz, which was held on May 3, 2009 and
encompassed over 37,000 acres from Carmel Valley south to middle Big Sur in Monterey
County. Partners in the regional sampling effort included the Monterey Regional Park
District, Big Sur Land Trust, White Rock Club, and several private landowners. The regional
SOD Blitz was advertised in a number of local media outlets and attracted 23 general public
participants from the Monterey Peninsula. A 1-hour workshop was held 2 weeks in advance
of the event, during which participants were taught to reliably identify California bay laurel
trees and SOD symptoms, and implement the simple sampling protocol. Sampling kits
containing a step-by-step sampling protocol, data sheet, printed aerial photo with
infrastructure layers, and 20 Ziploc® sandwich bags for sampled California bay laurel leaves
were distributed to workshop participants, in addition to educational information and a
complimentary phytosanitation hiking footwear bag. As with results for the second annual
Preserve-based SOD event, results for the regional event are expected in September 2009 and
will be made available via the Conservancy’s website at: http://www.slconservancy.org.
Through the development and implementation of the SOD Blitz events, the Santa Lucia
Conservancy, in collaboration with Dr. Matteo Garbelotto and local conservation partners, has
made significant strides in advancing SOD awareness as a community-based dynamic as well
as a private property concern.
The SOD Blitz events have spurred additional interest in other aspects of SOD management
on the Preserve, such as phytosanitation, prevention and treatment, and research. As a result,
the Conservancy has installed seasonal sanitation stations at the heads of frequently used
trails, host treatment workshops, and sponsors field research advancing the frontiers of our
ecological knowledge of P. ramorum and its life cycle in coastal forests. Additionally, the
Conservancy has developed the State’s first SOD management plan for private property,
entitled, “Sudden Oak Death on the Santa Lucia Preserve: A Community Approach,” which is
available upon request.
As with the management of other non-native pests such as plants, long-term datasets are
essential in tracking changes in the distribution and severity of bioinvasions so that
predictions may be made about their containment and treatment. The SOD Blitz events
provide a low-cost, reliable, educational, and predictable source of occurrence data with
which to monitor the spread and life cycle of P. ramorum in California’s coastal forests.
346
Proceedings of the Sudden Oak Death Fourth Science Symposium
Ambrosia Beetles and Their Associated
Fungi Appear to Accelerate Mortality in
Phytophthora ramorum-Infected Coast Live
Oaks1
Brice A. McPherson,2 David L. Wood,2 Nadir Erbilgin,3 and
Pierluigi Bonello4
Abstract
Infection of coast live oak (Quercus agrifolia) by Phytophthora ramorum, cause of sudden
oak death (SOD), is consistently followed by bark and ambrosia beetle attacks on the bark
overlying cankers. These beetles do not typically attack asymptomatic trees exhibiting healthy
green crowns. Beetle attacks reduced median survival of infected coast live oaks by 65 to 80
percent compared with beetle-free trees (McPherson and others, these proceedings). This
study was designed to explore the role of beetles and fungi in SOD and to determine the
sequential appearance of different fungal species.
We inoculated coast live oaks with P. ramorum in March 2005 at two sites in Marin County,
California, and then cut groups of infected trees at 6-month intervals thereafter.
Asymptomatic trees were also felled and treated as controls. Bolts were cut from logs and
dissected in the laboratory. From sections taken through these bolts at 15-cm intervals,
surface-sterilized wood samples were cultured on three growth media. We separated and
purified distinct morphotypes and amplified the internal transcribed spacer region (ITS) of the
rDNA operon. Amplicons were sequenced and blasted in GenBank. Only isolates showing
>95 percent ITS identity are reported here.
The diversity of fungal taxa isolated increased with time after inoculation and was
considerably greater in trees that had been attacked by ambrosia beetles. Several of these
fungi are known to be associated with Quercus spp., including Arthrographis cuboidea,
Botryosphaeria corticola, Pezicula cinnamomea, and Geosmithia langdonii. Other fungal
taxa, including Botryosphaeria sarmentorum, Kabatiella microsticta, Stereum hirsutum,
Trametes versicolor, and Truncatella angustata, have been reported to be pathogenic to
various hardwood species.
Beetles were reared from colonized trees, as well as dissected from their galleries, and were
cultured on the same media. From the bark beetle Pseudopityophthorus pubipennis, we
isolated Botryosphaeria corticola, Geosmithia pallida, Hypocrea schweinitzii, Mucor
racemosus, and two molds. Fungi were isolated from four ambrosia beetle species: Xyleborus
californicus (Hypocrea lixii, and two molds), Monarthrum scutellare (Hypocrea viridescens),
Monarthrum dentigerum (Lecanicillium cf. Psalliotae), and Xyleborinus saxeseni
(Pestalotiopsis sp.).
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
Department of Environmental Science, Policy, and Management, University of California, Berkeley,
CA 94720.
3
Department of Renewable Resources, University of Alberta, Edmonton, AB T6G 2E3, Canada.
4
Department of Plant Pathology, Ohio State University, Columbus, OH 43210.
Corresponding author bmcpherson@berkeley.edu.
347
GENERAL TECHNICAL REPORT PSW-GTR-229
Geosmithia spp. are found in association with bark beetles, including some that colonize
Quercus spp. in Europe (Kolarik and others 2005). An undescribed Geosmithia species has
been recently implicated in “thousand cankers disease,” an apparently introduced beetlevectored pathogen of walnuts in the western United States (Freeland and others 2009). The
significance of the association of a closely related Geosmithia species with P. ramoruminfected coast live oaks is unknown.
Literature Cited
Freeland, E., Kolarik, M., Utley, C., Cranshaw, W., and Tisserat, N. 2000. The
Geosmithia causing thousand cankers disease of walnut is a new species. Phytopathology 99
(6): S 37.
Kolarik, M.; Kubatova, A.; Cepicka, I.; Pazoutova, S. and Srutka, P. 2005. A complex of
three new white-spored, sympatric, and host range limited Geosmithia species. Mycological
Research. 109: 1323–1336.
348
Proceedings of the Sudden Oak Death Fourth Science Symposium
Mapping the Distribution of Sudden Oak
Death at the Urban-Wildland Interface:
Tilden Park in Berkeley, California1
Brice A. McPherson,2 David L. Wood,2 Maggi Kelly,2 and Gregory
Biging2
Abstract
In coastal California, sudden oak death is primarily a disease of forests. Prior to 2001, the
Phytophthora ramorum epidemic had not been found in the East Bay. This area east of San
Francisco Bay, though heavily urbanized, has extensive forested parklands and protected
watersheds that are similar in vegetation and climate to the heavily infested parts of Marin
County. Following several years of isolated reports of infected coast live oaks (Quercus
agrifolia) and California bay laurels (Umbellularia californica) in the East Bay, an extensive
outbreak was discovered in Tilden Park, adjacent to Berkeley, in October 2006.
Tilden Park lies at an urban-wildland interface, in an area that has been the site of several
major wildfires in the past century. As development continues to encroach on forested land,
pressures increase on these resources. In collaboration with the East Bay Regional Park
District, we are using the point-centered-quarter population density estimation method to
determine infection levels in a geographic information systems (GIS) context and to map the
distribution and variation in infection levels in coast live oaks within this park. This work is
part of a larger project to assess the extent of the epidemic in the regional parks and to
identify local factors influencing this distribution. A further goal of the study is to project the
forest composition and structure resulting from the predicted loss of coast live oaks.
Preliminary results indicate that coast live oak infections caused by P. ramorum are
distributed very unevenly across the park landscape. Individual coast live oak stands in the
area near the outbreak that was identified in 2006 have infection levels between 22 percent
and 61 percent. These levels are comparable to those observed in Marin County research sites
in 2000 (McPherson and others 2005). Other coast live oak stands in the park are currently
showing negligible infection levels.
Nearly two centuries of shifting land use patterns have created a patchwork of native
vegetation and introduced species, interspersed with buildings, recreational areas, and roads.
This heterogeneous landscape presents opportunities to understand the limits of the epidemic
in the context of multiple factors that influence its local abundance and expansion. The
resulting distributional maps will be designed to provide information to the land managers.
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
Dept. of Environmental Science, Policy, and Management, University of California, Berkeley, CA
94720.
Corresponding author: bmcpherson@berkeley.edu.
349
GENERAL TECHNICAL REPORT PSW-GTR-229
Literature Cited
McPherson, B.A.; Mori, S.M.; Wood, D.L.; Storer, A.J.; Svihra, P.; Kelly, N.M. and
Standiford, R.B. 2005. Sudden oak death in California: disease progression in oaks and
tanoaks. Forest Ecology and Management. 213: 71–89.
350
Proceedings of the Sudden Oak Death Fourth Science Symposium
Mapping the Impacts of Sudden Oak Death
Tree Mortality on Severity of the Big Sur
Basin Complex Fire1
Qingmin Meng,2 Sanjay Lamsal,2 Emily Holland,2 Douglas
Shoemaker,2 Margaret Metz,3 Kerri Frangioso,3 David M. Rizzo,3
and Ross K. Meentemeyer2
Introduction
Recent disturbance events, both biological and physical, continue to shape the
heterogeneous landscapes of California’s Big Sur ecoregion. Over the past decade,
Big Sur has experienced substantial mortality of oak (Quercus spp.) and tanoak
(Lithocarpus densiflorus) trees due to the emerging forest disease sudden oak death
(SOD). In 2008, a series of large wildfires burned a substantial portion of the region,
including a plot network established to study the spread and impacts of SOD.
Spatially-explicit maps of pre-fire tree mortality (Meentemeyer and others 2008) and
the existence of pre- and post-fire plot data provided an ideal opportunity to examine
feedbacks between landscape heterogeneity, mortality-related fuel loads, and fire
severity.
To examine landscape-level impacts of SOD on fire severity, we acquired
hyperspectral MASTER (MODIS/ASTER) and AVIRIS remote sensing data of the
burned area immediately following suppression of the wildfires. Of the 122
monitoring plots located within the burn perimeter, we quantified fire severity in 61
of the plots using BAER assessment methods prior to fall rains. Around each plot, we
also measured pre-fire landscape heterogeneity of vegetation type, tree mortality,
topography, and weather factors at the time of fire. These data were integrated with
our hyperspectral imagery to upscale plot-level burn indices to regional maps of fire
severity and to quantify the large-scale contribution of sudden oak death tree
mortality to fire severity.
Methodology
Multivariate analysis methods such as multivariate correlation, principal component
regression (PCR), and partial least squares regression (PLSR) have been applied in a
wide range of fields including environmental science, natural resources, ecology, and
geography. The main reason is that they have been designed to confront the situation
that there are many, possibly correlated, predictor variables, and relatively few
samples. This typically is a common situation in ecology, especially landscape
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
Department of Geography and Earth Sciences, University of North Carolina, Charlotte, NC 28223.
3
Department of Plant Pathology, University of California, Davis, CA 95616.
Corresponding author: qmeng1@uncc.edu.
351
GENERAL TECHNICAL REPORT PSW-GTR-229
epidemiology and fire ecology, in which extensive and on-time field surveys are time
consuming and people often do not have enough time and financial support to
complete large samples.
Principle component regression (PCR) and partial least squares regression (PLSR)
provide the potential approaches which can explore the maximum meaningful
information in hypespectral images and input the extracted information into fire burn
severity prediction and mapping. Another advantage of PCR and PLSR is that a
number of response variables (for example, burn severity of different forest stands’
layers) can be modeled or predicted at the same time.
Results
Using the hyperspectral images MASTER, we regionally predicted and mapped fire
severity of dominant tree layer, intermediate-sized tree layer, shrub layer, herb layer,
and composite burn index across the mixed oak and redwood-tanoak forests in Big
Sur, California. Our predictions for dominant tree layers are moderate strength, but
predictions for intermediate-sized tree layer and the overall burn index (CBI) are
relatively weak.
There are significant differences of fire severity across the Big Sur Complex Fire. For
example, multivariate analyses indicated that differences of fire severity between
mixed oak and redwood-tanoak forests are significant at the 0.005 level. After
accounting for topographic and vegetation community type, SOD mortality mapped
by Meentemeyer and others (2008) was positively correlated with fire severity.
Literature Cited
Meentemeyer, R.K.; Rank, N.E.; Shoemaker, D.; Oneal, C. and Rizzo, D.M. 2008.
Impacts of sudden oak death on tree mortality in the Big Sur ecoregion of California.
Biological Invasions. 10: 1243–1255.
352
Proceedings of the Sudden Oak Death Fourth Science Symposium
Phytophthora ramorum in USA Streams
From the National Early Detection Survey
of Forests1
Steven W. Oak,2 Jaesoon Hwang,3 Steven N. Jeffers,3
and Borys M. Tkacz4
Abstract
The National Phytophthora ramorum Early Detection Survey of Forests used terrestrial
vegetation survey protocols from 2003 to 2006. The pathogen was detected in only two out of
12,699 symptomatic plant samples collected and diagnosed in 39 states. Stream surveys
utilizing rhododendron leaf baits had been used successfully for P. ramorum early detection
and monitoring in California and Oregon forests since at least 2004 and were examined as an
alternative survey method in an effort to improve detection efficiency. The assumption that
stream survey by baiting is more efficient than terrestrial survey was supported by the
detection of the pathogen from 6 to 25 km downstream from the nearest known forest
infestations even before symptoms were detectable in low-altitude aerial surveys. Successful
pilot testing of stream baiting survey protocols in 11 states during 2006 resulted in full
implementation in 2007, and stream baiting surveys have continued to the present.
Stream baiting surveys were conducted by cooperators in state agencies and universities. The
number of streams surveyed in each state depended on risk as determined by host type,
climate, and potential for P. ramorum introduction. In the pilot survey year, four nonwounded rhododendron leaves in a mesh bag were deployed in each stream at monthly
intervals for five months during the growing season (May to September); leaves were exposed
for 1 or 2 weeks each month, depending on symptom development. After retrieval, leaves
were washed under running tap water and blotted dry, and then small pieces of water soaked
or discolored leaf tissue were removed and assayed for Phytophthora species and P. ramorum
by two methods—PCR and isolation on selective medium. Methods were modified slightly in
later years. In 2007, two bags of bait leaves were deployed in each stream during each baiting
period to compensate for occasional losses and to provide sufficient material for diagnostics.
In 2008, a sixth baiting period was added with direction to avoid deployment in mid-summer
months in states where high temperatures were presumed to be less favorable for growth and
sporulation of P. ramorum. Results are presented for the period from 2006 through the first
half of 2009.
Stream baiting surveys were completed in 320 unique streams in 28 states from 2006 through
2008 (table 1) with P. ramorum detected in waterways outside of known infested West Coast
counties each year. The first detection was reported during the first baiting period of the 2006
pilot survey year in a seasonal stream draining a positive nursery in Pierce County,
Washington. In 2007, this find was followed by a new detection in a river in King County,
Washington with multiple positive nurseries in the watershed and in a ditch and creek outside
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
USDA Forest Service, FHP, Southern Region, 200 WT Weaver Blvd., Asheville, NC 28804.
3
Department of Entomology, Soils, and Plant Sciences, 120 Long Hall, Clemson University, Clemson,
SC 29634.
4
USDA Forest Service, FHP, Washington Office, 1601 N Kent St RPC, Arlington, VA 22209.
Corresponding author: soak@fs.fed.us.
353
GENERAL TECHNICAL REPORT PSW-GTR-229
a positive nursery in Rankin County, Mississippi. Seven more new detections were reported
during 2008. Previously non-infested watersheds in the P. ramorum-endemic California
counties of Mendocino (three streams) and Humboldt (one stream) and Curry County, Oregon
(one stream), were found positive as well as waterways outside positive nurseries in Shelby
County, Alabama and Gadsden County, Florida (one stream each). In the first half of the 2009
survey season, new detections have been reported in waterways outside positive nurseries in
Montgomery County, Alabama and in Georgia (one each).
Table 1—Stream survey totals by region and year of survey
Number of streams
Year of survey
Number of
surveyed
Region of USA
states
2006 2007 2008
Total
Uniquea
West Coast
3
37
32
39
108
75
South
10
33
64
71
168
137
North Central
6
0
20
15
35
28
Northeast
9
24
37
29
90
80
National Total
28
94
153
154
401
320
a
Some streams were surveyed in multiple years.
In total, 12 first detections of P. ramorum have been made in seven states by the National
Early Detection Survey of Forests in 3.5 years using rhododendron leaf baiting of waterways.
Of these, five were made in streams draining watersheds near or adjacent to known positive
watersheds within the known range of P. ramorum and sudden oak death in West Coast
forests. The remaining seven streams are outside the known range but drain watersheds with
one or more confirmed positive woody ornamental crop nurseries (table 2). To date,
implementation of the U.S. Department of Agriculture, Animal and Plant Health Inspection
Service Confirmed Nursery Protocols in the infested nurseries have not prevented P. ramorum
from escaping nursery environments, as the pathogen has been found repeatedly in waterways
outside of these nurseries each year after initial detection and regulatory action.
Table 2—Location and year of P. ramorum detections outside the known West
Coast range of sudden oak death, January 2006 through June 2009
Location
Pierce Co., WA
King Co., WA
Rankin Co., MS
Gadsden Co., FL
Shelby Co., AL
GA
Montgomery Co., AL
a
2006
X
Year of detection
2007
2008
Xa
X
X
X
X
X
X
X
Jan-Jun 2009
X
X
X
X
X
X
X
Detection made by Washington State Department of Agriculture in a bait station not part of the
National Early Detection Survey of forests.
While asymptomatic plants, contaminated soil, and/or contaminated water inside confirmed
positive nurseries are strongly implicated as the sources of inoculum detected in these
streams, the possibility exists that exterior sources also are present. Repeated intensive
vegetation surveys in the environs of positive streams in Washington, Mississippi, Alabama,
Florida, and Georgia resulted in positive diagnoses of P. ramorum only in Mississippi during
winter 2007-2008, when four PCR-positive results were obtained from samples of three
different host and associated plant genera collected on two different survey dates. Stream
surveys using rhododendron leaf baits will continue, and intensive vegetation surveys in the
environs of positive streams will be repeated in order to ensure that P. ramorum has not
become established in natural ecosystems outside the currently known range of this pathogen.
354
Proceedings of the Sudden Oak Death Fourth Science Symposium
The Phytophthora Online Course: Training
for Nursery Growers1
Jennifer L. Parke,2 Jay Pscheidt,3 Richard Regan,4 Jan Hedberg,5
and Niklaus Grünwald6
Abstract
Oregon State University Extended Campus (Ecampus) launched an online training course for
the management of Phytophthora in nurseries. The course was developed with funds from the
U.S. Department of Agriculture’s Natural Resource Conservation Service through a grant to
the Oregon Department of Agriculture. To access the Phytophthora Online Course: Training
for Nursery Growers, visit http://ecampus.oregonstate.edu/phytophthora. English and Spanish
language versions are available.
The Phytophthora Online Course: Training for Nursery Growers provides access to all levels
of nursery personnel and the public worldwide. This free, non-credit course includes three
modules: biology, symptoms, and diagnosis; disease management; and Phytophthora
ramorum, the quarantine pathogen that causes sudden oak death in forest trees as well as
ramorum blight on nursery plants. It takes about 4 hours to complete the course. For an
optional $100 fee, nursery growers can earn a Certificate of Mastery after successfully
completing an online exam. Those who pass the exam can also earn four pesticide
recertification credits from the Oregon Department of Agriculture.
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
Department of Crop and Soil Science, Agriculture and Life Sciences Building 3017, Oregon State
University, Corvallis, OR 97331.
3
Department of Botany and Plant Pathology, Oregon State University, Corvallis, OR 97331.
4
Department of Horticulture, Oregon State University, Corvallis, OR 97331.
5
Oregon Department of Agriculture, 635 Capitol St NE, Salem, OR 97301.
6
USDA-ARS Horticultural Crops Research Laboratory, Corvallis, OR 97330.
Corresponding author: Jennifer.Parke@oregonstate.edu.
355
GENERAL TECHNICAL REPORT PSW-GTR-229
Establishment of Phytophthora ramorum
Depends on Woodland Diversity and
Species Composition1
Nathan Rank,2 Hall Cushman,2 Ross Meentemeyer,3 and David
Rizzo4
Abstract
Recent ecological theory predicts that disease establishment depends on local diversity and on
abundance of key vector species. These predictions have not been tested for plant pathogens
in woodland environments. Here we investigate the relationship between woody species
richness and the abundance of key tree species on establishment of Phytophthora ramorum on
its main foliar host, California bay laurel (Umbellularia californica). We also document
disease progression on bay laurel and coast live oak (Quercus agrifolia) over a 6-year period.
Oak woodlands of coastal California occur across a range of topographic and environmental
conditions that vary in local climate and support different species of woody plants. This
variation in microclimate and vegetation type yields high variability in the likelihood of
establishment of P. ramorum in California forests. For example, some areas support high
densities of bay laurel, the most important foliar host of P. ramorum, while this host is rare in
other areas. The risk of infection of coast live oak depends partly on the local abundance of
bay laurel, which serves as a source of inoculum and transmission to canker hosts.
We analyzed plant community data in 200 randomly located 15 x 15 m plots in a 275 km2
region in eastern Sonoma County, California. Within this network, bay laurel was the most
widely distributed woody species, occurring in 97 percent of plots, followed by coast live oak
(72 percent), Douglas-fir (Pseudotsuga menziesii) (47 percent), California black oak (Q.
kelloggii), (45 percent), Pacific madrone (Arbutus menziesii) (43 percent), Oregon white oak
(Q. garryana) (43 percent), and toyon (Photinia arbutifolia) (41 percent). When plots were
established in 2003, we measured over and understory abundance of woody species and
installed microclimate loggers to measure understory temperature and relative humidity. Since
2003, we have conducted annual surveys of forest structure for each main woody host of P.
ramorum. In addition, we surveyed disease severity of P. ramorum from 2004 to 2009
through timed counts of symptomatic leaves on bay laurel. Using data from this plot network,
we asked the following questions: 1) Does woody species richness affect pathogen prevalence
on bay laurel hosts? 2) Does pathogen prevalence on bay laurel depend on the presence of
coast live oak and black oak? 3) How does disease establishment on bay laurel, coast live oak,
and black oak progress over time? 4) Does pathogen abundance on bay laurel relate to
infection rate of coast live oak?
To assess the relationship between woodland species richness and P. ramorum establishment,
we conducted multiple regressions that included number of species and bay laurel stem
number as independent variables and either total number of bay laurel symptomatic leaves per
plot or mean number of symptomatic leaves per bay laurel stem as dependent variables. The
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
Sonoma State University, 224 Darwin Hall, 1801 East Cotati Ave., Rohnert Park, CA 94928.
3
University of North Carolina at Charlotte, Charlotte, NC 28223.
4
Department of Plant Pathology, University of California at Davis, Davis, CA 95616.
Corresponding author: rank@sonoma.edu.
356
Proceedings of the Sudden Oak Death Fourth Science Symposium
total number of leaves provided an index of inoculum load and the mean number of leaves per
stem provided an index of disease prevalence. The multiple regressions revealed that
pathogen abundance and prevalence declined with woodland species richness and increased
with bay laurel abundance (P < 0.005 for both comparisons). This relationship suggests that
species diversity acts as an encounter reduction mechanism with a dilution effect on disease
risk.
Our results also showed that disease severity was lower in plots where coast live oak was
present than in plots where it was absent (two-way factorial ANCOVA with bay laurel stem
number as covariate; P < 0.01 for both comparisons). No such effect was observed for black
oak. Bay laurel co-occurs with coast live oak more often than any other pair of woody species
in our plot network (52 percent of plots) and the number of bay laurel stems per plot is
negatively related to the number of coast live oak stems (Spearman’s r = -0.14, P < 0.05). The
effect of coast live oak on P. ramorum establishment suggests that pathogen transmission is
reduced in areas where bay canopy is reduced by competition with coast live oak. Coast live
oak seems to be one of the most important woodland species influencing pathogen
establishment on bay laurel, and as the oak canopy declines due to disease progression, we
may observe an increase in pathogen abundance and prevalence on bay laurel, which will
eventually result in even greater levels of disease on coast live oak.
We also quantified disease progression on bay laurel and coast live oak from 2004 to 2009.
From the outset, symptoms of P. ramorum have been observed on the vast majority of bay
laurel stems (> 75 percent). Between 2005 and 2006, the number of symptomatic stems
increased to over 90 percent of all bay laurel individuals, and it leveled off after 2007. In
contrast, relatively few coast live oak or black oak stems show symptoms of P. ramorum
infection (assessed by the presence of a canker on the trunk). In 2004, 3 percent of coast live
oak stems possessed cankers, but the prevalence increased to 21 percent by 2009. The largest
year to year increase in proportion of oak stems with cankers was observed from 2006 to
2007, suggesting a 1 year lag between the changes in the proportion of symptomatic bay
stems and disease progression on oak. By 2009, we observed mortality of 5.8 percent of the
total number of coast live oak stems (n = 742) after observation of a P. ramorum canker on
the stem. This number represents a lower bound estimate of P. ramorum-related mortality.
Our data support the general hypothesis that oak mortality increases with increasing amounts
of inoculum on leaves of bay laurel. The probability that a coast live oak stem possessed a
canker by 2009 increased with increasing pathogen abundance (quantified by bay laurel
symptomatic leaf number; logistic regression weighted by oak stem number; n = 128 plots, P
< 0.001). We also observed that the probability of coast live oak mortality in a study plot by
2009 increased with increasing pathogen abundance on bay laurel (logistic regression
weighted by oak stem number; n = 128 plots, P < 0.001). As coast live oak infection and
mortality progress further, we may observe the same relationships between woodland species
diversity and disease progression in coast live oak that we documented for bay laurel.
357
GENERAL TECHNICAL REPORT PSW-GTR-229
Using Rain Bucket Spore Traps to Monitor
Spore Release During SOD Eradication
Treatments in Oregon Tanoak Forests1
Paul Reeser,2 Everett Hansen,2 Wendy Sutton,2 Alan Kanaskie,3
Jon Laine,3 Michael Thompson,3 and Ellen Michaels Goheen4
Introduction
The complete Oregon eradication treatment for Phytophthora ramorum, the cause of
sudden oak death, calls for early detection and: 1. prompt killing of infected trees
plus a buffer of visibly healthy trees using herbicide when possible to prevent resprouting; 2. falling the killed trees and cutting other host plants; and 3. burning all
slash. The goal is to halt sporulation and dispersal of sporangia as quickly as
possible. These treatments are expensive, slow, and resisted by some landowners.
There has been no rigorous comparison of the effectiveness of these tree killing
methods on inoculum production, however. We report on trials to test these treatment
variables, using baited rain traps to monitor production of sporangia under the
different eradication treatment conditions.
Methods
Rain traps baited with rhododendron (Rhododendron macrophyllum) and tanoak
(Lithocarpus densiflorus) leaves were placed beneath tanoak trees in areas distant
from known infection, in known infested stands before eradication treatments began,
and in infested stands during and after eradication treatments. Traps were moved as
the treatments progressed, and traps were placed in new areas as P. ramorum was
confirmed. Stands of green, apparently healthy tanoak were used as negative
controls. Trapping continued for 18 months, ending April 2008.
Two-gallon white HDPE plastic buckets were lined with thin plastic bags. Bait leaves
were placed in the bucket along with approximately 375 ml de-ionized water. The
bucket was covered with a screen to keep out large debris. At retrieval (after 2 weeks
of exposure), water depth was measured, the plastic liner replaced, and new bait
leaves were added.
Bait leaves were kept in Ziploc® bags in a cooler and transported to the laboratory,
where they were washed in tap water and blotted dry. Necrotic areas of leaves were
plated in semi-selective CARP+ medium (Corn meal agar base amended with 30 ppm
Benlate 75WP, 10 ppm Na-natamycin, 200 ppm Na-ampicillin, 10 ppm rifamycin
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
Dept. Botany and Plant Pathology, Oregon State University, Corvallis OR 97331.
3
Oregon Dept. of Forestry, 2600 State Street, Salem, OR 97310.
4
USDA Forest Service, SW OR Forest Insect and Disease Service Center, Central Point OR 97502.
Corresponding author: akanaskie@odf.state.or.us.
358
Proceedings of the Sudden Oak Death Fourth Science Symposium
SV, and 25 ppm hymexazol). Petiole ends were always included whether necrotic or
not. Isolation plates were incubated in the dark at 20 ºC for 7 to 10 days, then
examined for the growth of P. ramorum, which was recognized by a combination of
distinctive hyphae, chlamydospores, and sporangia.
Trap placements were classified by site infestation status (infested or healthy),
eradication treatment progress (untreated, partially treated, treatment completed), and
canopy condition (infected tanoak, brush pile, non-host species, and others). There
were usually 5, 10, or 15 traps placed at a site, depending on size of the tanoak stand
or eradication area. Traps were scored as positive or negative for P. ramorum for
each sampling interval based on culture results.
Results
Phytophthora ramorum inoculum is not generally distributed in the Curry County
quarantine area. The pathogen was seldom recovered (1 percent of traps) from
apparently healthy stands away from known infested areas. P. ramorum was
recovered more often from traps placed inside the eradication area (29 percent of
traps) than in traps placed in the eradication area perimeter, around 300 feet from the
nearest infected tanoak (less than 1 percent of traps).
Recovery of P. ramorum was reduced in traps placed in infested stands during and
after site treatment (table 1).
Table 1—Recovery of P. ramorum in infested stands
Proportion traps
Site condition
Treatment status
positive for P.
ramorum
Infested
None
0.60
Infested
Partial
0.27
Infested
Completed
0.11
Healthy
None
0.01
Number of traps
447
301
1473
343
In infested tanoak stands with active sporulation, the probability of detecting P.
ramorum varied with canopy condition (table 2). Recovery of P. ramorum was
reduced after application of the hack and squirt herbicide (imazapyr) treatment, but
this effect was often months in occurring. Spores were recovered from trees with
brown foliage due to hack and squirt on some occasions. In at least one location, P.
ramorum was detected in open ground at a fully treated site 10 months after all
tanoak canopy had been removed, piled, and burned. This may have been due to a
blow-in event or rain splash from the forest floor.
359
GENERAL TECHNICAL REPORT PSW-GTR-229
Table 2—Recovery of P. ramorum in relation to crown condition
Proportion traps
Canopy condition in
positive for P.
Number of traps
eradication area
ramorum
Symptomless Tanoak
0.31
71
Infected Tanoak
0.70
371
Tanoak - Herbicide
0.53
156
Myrtlewood
0.41
129
Brush Pile
0.21
189
Non-Tanoak Species
0.15
185
Alder
0.05
86
Rhododendron
0.0
23
Huckleberry
0.0
25
Open Ground
0.48
27
In infested stands, P. ramorum was recovered at a high frequency, apparently
correlated with rainfall, as measured by water depth in the rain trap at the time of
retrieval (fig. 1). In rainy periods, the pathogen was recovered from 60 to 95 percent
of the rain traps. This effect may be confounded with season and changing site
characteristics as eradication proceeded.
infected partial treated
infected treated
rainfall
30
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
20
15
10
5
0
Fall
Winter
Spring
Summer
Fall
Winter Spring
Figure 1—Inoculum was available during rain events throughout the year. P.
ramorum was recovered from bait leaves in rain traps in 34 of 36 2-week sample
periods. Periods with reduced rainfall (Red Mound Remote Automated Weather
Station, http://www.wrcc.dri.edu/cgi-bin/rawMAIN.pl?orOREM) tended to have
reduced detection of P. ramorum.
Summary
1. Inoculum is available during rain events throughout the year. P. ramorum was
recovered from bait leaves in rain traps in 34 of 36 2-week sample periods.
360
Rainfall (cm)
25
10/31/2006
11/9/2006
11/13/2006
11/29/2006
12/12/2006
1/3/2007
1/16/2007
2/6/2007
2/20/2007
3/7/2007
3/21/2007
4/4/2007
4/18/2007
5/1/2007
5/15/2007
6/5/2007
6/19/2007
7/9/2007
7/23/2007
8/23/2007
9/5/2007
9/19/2007
10/2/2007
10/16/2007
10/30/2007
11/13/2007
11/27/2007
12/12/2007
1/8/2008
1/22/2008
2/5/2008
2/20/2008
3/5/2008
3/19/2008
4/9/2008
4/23/2008
Proportion positive traps
infected untreated
Proceedings of the Sudden Oak Death Fourth Science Symposium
2. Phytophthora ramorum inoculum is not generally distributed in the Curry County
quarantine area. P. ramorum was seldom recovered from apparently healthy stands
away from known infested areas.
3. In infested stands, P. ramorum was recovered at a high frequency, apparently
correlated with rainfall. In rainy periods, the pathogen was recovered from 60 to 95
percent of the rain traps.
4 Recovery in infested stands during and after site treatment was reduced and the
pathogen was never recovered from traps placed just outside the treatment areas.
A second trial is underway, with buckets placed beneath individual trees of known
infection and treatment status. This will allow us to separate stages of the treatment
process and better assess which treatments have the biggest impact on inoculum
production.
Acknowledgments
Thanks to the U.S. Department of Agriculture, Forest Service (USDA FS), Pacific Northwest
Region (Region 6) and USDA FS Pacific Southwest Research Station, for funding.
361
GENERAL TECHNICAL REPORT PSW-GTR-229
Chip-Based On-Site Diagnosis of
Phytophthora ramorum1
M. Riedel,2 S. Julich,2 M. Kielpinski,2 R. Möller,2 W. Fritsche,2
S. Wagner,2 T. Henkel,2 A. Breitenstein,2 and S. Werres2
Abstract
The pathogen Phytophthora ramorum Werres, De Cock, & Man in’t Veld which attacks a
wide range of host plants and is a worldwide threat to natural ecosystems and the nursery
industry. To prevent the spread of this pathogen with latently infected plants, easy to handle
diagnostic systems with high sensitivity and high throughput are demanded. Furthermore,
such methods should give results within a short time on site of inspection. A 3-year project is
underway to create a chip-formatted PCR system combined with a chip-based electrical
microarray that will be adapted for the detection of P. ramorum in plant tissue. A stationary
PCR chip with integrated microstructured heaters and temperature sensors has been developed
for the amplification of specific Phytophthora DNA fragments. A microfluidic system
connects the PCR chip with a microarray where the labelled DNA fragments are detected. The
miniaturization of the chip formatted PCR and array system enables high portability, low
input of energy, smaller amounts of expensive analytic chemicals and due to low reaction
volumes very fast reaction times.
Introduction
Determination of Phytophthora ramorum in latently infected plants or in plants with
unspecific symptoms presupposes a (mostly centralized) microbiological laboratory,
experienced personnel, and time. But, to prevent the spread of P. ramorum with
latently infected plants, decisions in phytosanitary management have to be made fast.
Therefore, easy to handle diagnostic systems with high throughput and high
sensitivity are in need.
Here we present the development of a portable chip-based nucleic acid detection
system for the identification of P. ramorum and other commercially important
Phytophthora species. This system will enable the fast and specific diagnosis of a
great number of samples in the field.
Material and Methods
For the demonstration of the detection system, the ITS2 area of the ribosomal DNA
was chosen as the target sequence for the identification of Phytophthora species. For
many Phytophthora species, this region is sufficient for discrimination. To optimize
the chip-formatted reactions, both parts of the detection process (PCR and
hybridization) were developed in separate approaches.
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
Julius Kühn Institut - Federal Research Centre for Cultivated Plants, Institute for Plant Protection in
Horticulture and Forests, Messeweg 11/12, D – 38104 Braunschweig, Germany.
Corresponding author: marko.riedel@jki.bund.de.
362
Proceedings of the Sudden Oak Death Fourth Science Symposium
PCR-Chip Module
A stationary PCR chip (fig. 1A) was developed which integrates microstructured thin
heater structures and temperature sensors. The PCR reaction is performed in a PDMS
(Polydimethylsiloxane) reaction chamber on top of the stationary PCR chip. The
reaction chamber is disposable (one time usage) so that any cross contaminations can
be prevented. The PCR protocol is designed as a two-step protocol (combined
annealing and elongation step). After PCR, the reaction products are transported to
the detection chip by an automatic microfluidic system.
DNA-Chip Module
The PCR products are analyzed with a chip formatted microarray (fig. 1B). This
disposable chip combines a microfluidic hybridization chamber with electrode
structures. The hybridization is designed as flow through hybridization. DNA probes
are spotted at the electrode gap. Detection of the hybridization with species-specific
probes is performed by enzymatic silver enhancement. Successful hybridization
results in an aggregation of silver particles at the electrode gap. The resulting change
of the electrical resistance is measured. The hybridization can also be analyzed
optically by measuring the gray scale value of the silver aggregation at the spotting
point.
Both chips are combined in a united platform for the amplification, labelling, and
qualitative detection of Phytophthora DNA. The software for control maintenance
and analysis of the reactions can be run on an average Personal Digital Assistant
(PDA).
A
B
Figure 1—Stationary PCR chip with four integrated microstructured heaters,
temperature sensors, and PDMS chamber for on-chip amplification and labeling of
DNA (A); gold microarray chip combined with a microfluidic PDMS channel
(prototype), specific oligonucleotide probes are spotted at 42 detection sites (B).
Preliminary Results
PCR protocols for chip-based amplification of the approximately 550bp ITS2
sequences were adapted and optimized. Amplification products produced with the
363
GENERAL TECHNICAL REPORT PSW-GTR-229
PCR chip are identical with PCR products from conventional thermocyclers (fig. 2).
The two-step PCR protocol helped to reduce the amplification time by more than 50
percent. Advantages of on-chip amplification are very short heating and cooling
times and reduced input of chemicals and energy. These advantages can be achieved
by the low reaction volume (up to 0.5 µl) of the chip PCR.
For the identification of P. ramorum, different species-specific probes (30 basepairs)
located at the ITS2 region were developed and tested in hybridization assays. The
specificity of the P. ramorum probes was tested with a range of closely related
Phytophthora species including P. lateralis and P. hibernalis (fig. 3). The flow
through hybridization in the microfluidic hybridization chamber has the advantage of
reducing the reaction time and the reagent volumes.
M
0
1
2
A
B
C
D
3
M
Figure 2—PCR chip:
amplification of the P. ramorum
ITS2 fragment.
0: negative control (buffer);
1: PCR in conventional
thermocycler;
2: PCR in PDMS chamber
chip;
3: PCR on glass chip;
M: DNA size marker.
Figure 3—DNA chip:
hybridization of P. ramorum-specific
probe with ITS2 PCR products (ca
550bp) of P. ramorum and closely
related Phytophthora taxa P. lateralis
and P. hibernalis (optical analysis of
gray scale value of the aggregated
silver nanoparticles at electrode gap).
A: specific hybridization: P. ramorum x
P. ramorum;
B: unspecific hybridization: P. ramorum
x P. lateralis;
C: unspecific hybridization: P. ramorum
x P. hibernalis;
D: negative control P. ramorum x
buffer.
Acknowledgments
This work is funded by the German Federal Agency for Agriculture and Food (BLE); (FKZ
28-1-42.027-06).
364
Proceedings of the Sudden Oak Death Fourth Science Symposium
Burn it, Chip it, or Tarp it, but Just Don’t
Move it: Managing Oak Firewood Infested
With the Goldspotted Oak Borer, Agrilus
coxalis auroguttatus1
Steven J. Seybold,2 Tom W. Coleman,3 and Mary Louise Flint4
Abstract
In June 2008, a new and potentially devastating pest of oaks, Quercus spp., was discovered in
San Diego County, California. This pest, the goldspotted oak borer (GSOB), Agrilus coxalis
auroguttatus Schaeffer (Coleoptera: Buprestidae), colonizes the sapwood surface and phloem
of the bole and larger branches of at least three species of Quercus. Larval feeding kills
patches and strips of the phloem and cambium resulting in crown dieback followed by
mortality. Since 2002, aerial surveys in San Diego County have detected about 20,000 dead
oaks. In a survey of forest stand conditions at three sites in this area, 67 percent of the oaks
had external or internal evidence of GSOB attack. The damage is worst for mid-story and
massive overstory oaks in the red oak group (subgenus Quercus, section Lobatae), which are
succumbing to colonization at a rate of 90% in some areas. In the worst cases, up to 50% of
oaks >12 cm are dying. Because of its recent discovery in California, specific management
practices for GSOB are still being tested. However, given its potential for damage, landscape
and land managers need guidelines for managing this pest now. We are beginning to develop
the first components of an integrated pest management (IPM) program for this new pest based
on IPM principles developed for other well-known Agrilus spp. pests of shade trees, for
example, the bronze birch borer, A. anxius, and the emerald ash borer, A. planipennis.
Key techniques in the plan include monitoring the adult flight period and sanitation of oak
firewood by various methods including solarization. Determining the flight period and
enhancing survey techniques will be crucial for managing GSOB populations and preventing
oak mortality. In 2008, purple and lime green flight-intercept panel traps were found to be
effective for trapping GSOB. We assessed the flight period from June to November in 2008
and resumed trapping in January 2009. Various lures (manuka oil, phoebe oil, and ethanol)
and trap placements were also assessed in 2009 to improve efficacy of trap catch and
detection.
Movement of firewood is a major pathway of dispersal for many insect pests in the U.S., and
represents one hypothesis for the introduction of GSOB into California. Outreach efforts to
minimize further movement of infested firewood within the state and to encourage treatment
of firewood are also essential. We assessed three solarization treatments of firewood: 1) direct
sun exposure; 2) tarping with clear plastic and sun exposure; and 3) caging of firewood and
storage in a shaded area to monitor normal woodborer emergence (control treatment). These
treatments were replicated at two elevations (1,220 and 1,740 m) at sites that include recently
1
A version of this poster was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
USDA Forest Service, Pacific Southwest Research Station, 720 Olive Dr., Suite D, Davis, CA 95616.
3
USDA Forest Service, Forest Health Protection, Southern California Shared Service Area, 602 S.
Tippecanoe Ave., San Bernardino, CA 92373.
4
University of California, Davis, Statewide IPM Program and Department of Entomology, Davis, CA
95616.
Corresponding author: sjseybold@gmail.com.
365
GENERAL TECHNICAL REPORT PSW-GTR-229
GSOB-killed coast live oaks (Quercus agrifolia) and California black oaks (Q. kelloggii),
respectively. We are also looking at the efficacy of chipping firewood.
This study will develop methods for reducing GSOB populations and oak mortality.
Educational materials from this study will be widely disseminated to the public to reduce the
movement of untreated firewood and slow the spread of GSOB throughout California.
Introduction
The goldspotted oak borer (GSOB), Agrilus coxalis auroguttatus Schaeffer
(Coleoptera: Buprestidae), is a new threat to oaks in southern California. Tree
mortality has been recorded since 2002, but was attributed to drought for many years.
This phloem/wood borer was first collected in San Diego County in 2004, but not
linked to the on-going tree mortality until June 2008 (Coleman and Seybold 2008a).
We hypothesize that GSOB was most likely introduced on firewood from its native
distribution in Arizona and Mexico. There are no collection records of GSOB from
localities between the native habitat and California, ruling out a natural range
expansion for GSOB (Coleman and Seybold 2009, 2010).
Aerial surveys of San Diego County have detected an estimated 20,000 dead oaks
over the past 8 years. In this area, GSOB injures and kills coast live oak (Quercus
agrifolia), California black oak (Q. kelloggii), and canyon live oak (Q. chrysolepis).
There is no previous record of injury or mortality to oaks from GSOB in Arizona and
Mexico. This species has been rarely collected in its native region and very little was
known about its life history and biology prior to 2008.
Signs of GSOB injury include thinning crowns, black staining of the outer bark
primarily along the bole, bark removal by woodpecker foraging, and D-shaped adult
exit holes (Coleman and Seybold 2008b). Repeated larval feeding at the interface of
the phloem and xylem kills patches of the cambium, which leads to tree mortality
after several years.
This new pest is killing high-value landscape and shade trees, impacting aesthetic
beauty of urban-wildland landscapes and property values. Loss of a single shade tree
or the cost of removing hazardous trees around dwellings entails significant costs
(upwards to $50,000). We describe a provisionary plan for an integrated pest
management (IPM) program for this new pest based on IPM principles developed for
other well-known Agrilus spp. pests of shade trees.
Key techniques in the plan include monitoring adult flight period and sanitation of
oak firewood by various methods including solarization. We focussed on three main
objectives for adult trapping: 1) determine the adult flight period with purple and
lime-green prism flight-intercept sticky panel traps; 2) determine the most effective
height for purple and lime-green prism traps; and 3) assess lures for enhancing adult
trap catch.
Movement of firewood is a major pathway of dispersal for many insect pests in the
U.S., and represents the primary hypothesis for the introduction of GSOB into
California. There is evidence of previous and current movement of firewood into and
out of the infested area of San Diego County. Outreach efforts to minimize further
movement of infested firewood within the state are also essential. We assessed three
366
Proceedings of the Sudden Oak Death Fourth Science Symposium
solarization treatments of firewood: 1) direct sun exposure; 2) tarping with clear
plastic and sun exposure; and 3) caging of firewood and storage in a shaded area to
monitor normal woodborer emergence (control treatment).
Methods and Results
Adult Trapping
In 2008, purple prism flight-intercept panel traps were established in late June in two
forest stands of coast live oak (four traps/site). Maximum flight was recorded during
the first week that the traps were in place and flight declined until it ceased in
November (Coleman and Seybold 2008a).
In January 2009, this work was continued with both purple and lime-green traps, but
was expanded to two elevations (1,035 and 1,765 m), which also represent two
distinct host distributions (coast live oak and California black oak). Three sites were
established at each elevation and three traps of each color were placed at each site.
The traps were placed at three heights: low (1.5 m), medium (3 m), and high (4.6 m).
Traps were monitored weekly until mid-September.
In 2009, GSOB did not begin flying until the week of May 15 to 22 (lower elevation
site). GSOB flight did not begin until the week of May 29 to June 5 at the high
elevation site. Peak flight activity occurred between mid-June and early July at both
sites. Flight activity declined rapidly in early August; beetles were still trapped in
mid-September, but the numbers were very low. Purple prism traps were more
effective at catching GSOB than lime-green traps, but the difference was not
significant. Purple prism traps were most effective when hung at 3.0 m above
ground, whereas lime-green traps were most effective when hung at 4.5 m above
ground.
From June to August 2009, four lures were assessed to enhance trap catch: 1) manuka
oil; 2) phoebe oil; 3) (Z)-3-hexenol; and 4) no lure. Twenty-four traps were used
across three sites (four green and four purple/site). Manuka oil is a steam distillate of
the New Zealand manuka tea tree, Leptospermum scoparium, whereas phoebe oil is a
steam distillate of Brazilian walnut, Phoebe porosa. None of the trap baits were
effective at attracting GSOB.
Firewood
Firewood pieces (~30 cm long x 10-20 cm dia.) from recently killed (<1 yr) coast
live oak and California black oak were collected in the winter of 2008-2009 from >15
trees and the pieces were randomized within species prior to assignment to three
treatment groups (i.e., six treatments total). The experiment was established in
February 2009 and duplicated at two sites: Low (1,220 m) and high (1,740)
elevations. Treatment replicates consisted of six to eight pieces of wood that were
caged in aluminum mesh screening. The three treatment groups were: 1) direct sun
exposure; 2) tarping with clear plastic and sun exposure; and 3) caging of firewood
and storage in a shaded area to monitor normal woodborer emergence (control
treatment). For the tarping treatment, two layers of thick, 8 mil clear plastic were
used. The tarp edges were covered with soil to prevent beetles from escaping and to
367
GENERAL TECHNICAL REPORT PSW-GTR-229
trap heat beneath the plastic. All treatments were sampled weekly for GSOB
emergence until September. HOBO data loggers recorded temperature and relative
humidity in selected replicates from all treatment groups. Mean GSOB emergence
pooled over oak species and elevation was the primary variable that we measured
(N=36). Treatments were analyzed statistically by using a mixed model of analysis
of variance.
Adult GSOB were recovered from firewood of both species of oaks. Adult
emergence began the week of May 22 and continued into July and August. No
significant differences were detected among the three treatments. Initial beetle
emergence occurred two weeks earlier in the direct sun exposure and tarping
treatments, most likely due to the increased temperatures in the sun-exposed wood
piles. However, GSOB emergence was significantly greater in piles of coast live oak
than in piles of California black oak.
Temperature and relative humidity were highest in tarped treatments when measured
over the duration of the study; the temperatures under the tarp reached 52.4 °C during
daytime highs. Temperatures were also slightly higher in direct sunlight treatments
than in shaded controls.
Discussion
Our capability for trapping GSOB adults is in its infancy. Purple-colored traps were
slightly more effective than lime-green-colored traps and a height of 3 m (relative to
1.5 or 4.6 m) was the most effective. None of the bait combinations that we tested
seemed to enhance trap catch and much more needs to be done to identify specific
host- or beetle-produced attractants for GSOB. The flight of GSOB appears to begin
in mid-May and then peak in mid-June to early July (Coleman and Seybold 2008a,
this study). Flight declines rapidly in August and then continues at a very low level
into the fall. Additional flight trapping data will allow us to better understand the
biology and phenology of this new pest. This information will be critical to
effectively time preventive pest management practices including insecticide
applications when they are necessary.
We reared GSOB in spring/summer 2009 in the field from firewood pieces that had
been cut from trees that died in 2008. New adult beetles emerged from this wood
from mid-May through mid-summer. Thus, the immature and mature stages of the
beetle can survive in small pieces of cut wood through the fall and winter, even under
harsh treatment conditions. This underscores and justifies the concern of moving and
properly managing firewood. Also this supports the hypothesis that GSOB was
introduced into California in firewood. Proper firewood management will potentially
slow the spread of this insect from San Diego County into other parts of California.
Although the direct sunlight and tarping treatments did not suppress GSOB
emergence relative to the shaded control, tarping firewood will likely prevent adults
from escaping. Future studies (2010) will assess other types of plastic, the capacity
of plastic to trap emerging beetles, and the effect of chipping as an option for
managing infested firewood. Chipping infested wood may be the best method for
reducing GSOB population density, and this technique has been worked out recently
for the emerald ash borer, A. planipennis, in the eastern U.S. Treating firewood may
368
Proceedings of the Sudden Oak Death Fourth Science Symposium
reduce localized beetle populations, which can protect adjacent high-value shade
trees.
Acknowledgments
The authors would like to thank the Pacific Southwest Region Forest Health Protection, the
Pacific Southwest Research Station, and the Cleveland National Forest, Descanso Ranger
District for logistical support of this work; Andreana J. Cipollone (U.S. Department of
Agriculture, Forest Service) and Stacy Hishinuma and Deguang Liu (Department of
Entomology, University of California, Davis) and Andrew D. Graves (Department of Plant
Pathology, University of California, Davis) for assistance with the project. Initial critical
funding for this work was provided primarily by the USDA Forest Service, Pacific Southwest
Research Station, Invasive Species Program; Forest Health Protection, Pacific Southwest
Region; and by the Forest Health Monitoring Program.
Literature Cited
Coleman, T.W. and Seybold, S.J. 2008a. Previously unrecorded damage to oak, Quercus
spp., in southern California by the goldspotted oak borer, Agrilus coxalis Waterhouse
(Coleoptera: Buprestidae). Pan_Pacific Entomologist. 84: 288–300.
Coleman, T.W. and Seybold, S.J. 2008b. New pest in California: The goldspotted oak
borer, Agrilus coxalis Waterhouse. USDA Forest Service, Pest Alert, R5-RP-022, October 28,
2008, 4 pp.
Coleman, T.W. and Seybold, S.J. 2009. Striking gold in southern California: Discovery of
the goldspotted oak borer and its central role in oak mortality. pp. 12-16. in McManus, K.A.
and Gottschalk, K.W. (eds.). Proceedings, 20th U.S. Department of Agriculture Interagency
Research Forum on Gypsy Moth and Other Invasive Species 2009, January 13-16, 2009; Gen.
Tech. Rep. NRS-P-51. 114 pp. Newtown Square, Pennsylvania: USDA, Forest Service,
Northern Research Station, December 2009.
Coleman, T.W. and Seybold, S.J. 2010. GSOB ≠ SOD: Tree mortality from the goldspotted
oak borer in oak woodlands of southern California. In Frankel, S.J.; Kliejunas, J.T.; Palmieri,
K.M. tech. cords. 2010. Proceedings of the Sudden Oak Death Fourth Science Symposium.
Albany, CA: Pacific Southwest Research Station, Forest Service, U.S. Department of
Agriculture. This volume.
369
GENERAL TECHNICAL REPORT PSW-GTR-229
Stream Monitoring for Detection of
Phytophthora ramorum in Oregon Tanoak
Forests1
Wendy Sutton,2 Everett Hansen,2 Paul Reeser,2 Alan Kanaskie,3
and Harvey Timeus3
Abstract
Stream monitoring using leaf baits for early detection of Phytophthora ramorum is an
important part of the Oregon sudden oak death (SOD) program. Fifty-eight streams in and
near the Oregon quarantine area in the southwest corner of the state are currently monitored.
The watersheds monitored range in size from 8 to 3592 ha, with a combined area of 323 656
ha. (fig. 1).
Figure 1—Map of southwest Oregon shows current quarantine area. Primary sites
have been continuously sampled over time; secondary sites have been sampled
varying lengths of time.
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
Department of Botany and Plant Pathology, Oregon State University, Corvallis, OR 97331.
3
Oregon Department of Forestry, 2600 State Street, Salem, OR 97310.
Corresponding author: hansene@science.oregonstate.edu.
370
Proceedings of the Sudden Oak Death Fourth Science Symposium
Rhododendron (Rhododendron macrophyllum) and tanoak (Lithocarpus densiflorus) leaf baits
in mesh bags are exchanged every 2 weeks throughout the year.
Early in the monitoring all samples were assayed by multiplex ITS PCR and isolation on
selective medium. The correlation between the two tests is strong and eventually only PCR
was performed (table 1). If a previously negative site assayed positive for P. ramorum, the
sample was plated on selective medium.
Table 1—Correspondence between culture and multiplex PCR results for
detection of P. ramorum in 1,804 stream bait samples collected from 2003
through 2007 and tested by both methods
Diagnostic Method
Culture
PCR
Percent of Samples
Positive
Positive
5.3
Positive
Negative
0.8
Negative
Positive
11.8
Negative
Negative
82.2
Once a site tested positive for P. ramorum, follow-up intensive ground surveys located
infected tanoaks or other host plants an average of 295 m upstream from the bait station.
Stream baits have been positive for P. ramorum up to 1150 m (stream distance) from the
probable inoculum source.
In eight watersheds P. ramorum was detected in stream baits before we found infected plants.
An example is stream WA65 which drains an old-growth redwood stand on federal land. It
was first baited for P. ramorum in May 2004, and was first culture and PCR positive in
October 2005. Despite earlier searches, infected plants were not located until April 2006; all
host plants on the sites were cut and burned later that year.
Phytophthora ramorum was detected by both culturing and multiplex PCR of leaf baits in all
seasons of the year. Sampling was sometimes impossible in late fall and winter due to
fluctuating water levels associated with winter storms. In streams with lower frequency of
recovery of P. ramorum, a seasonal trend was evident. Detection of P. ramorum was usually
lowest from December through April, with peak recovery rates from July through October
(fig. 2).
P ercent P CR P ositive
30
25
20
15
10
5
M ar
Apr
M ay
Jun
Jul
Aug
S ep
Oct
Nov
Dec
Jan
Feb
M ar
Apr
M ay
Jun
Jul
Aug
S ep
Oct
Nov
Dec
Jan
Feb
M ar
Apr
M ay
Jun
Jul
Aug
S ep
Oct
Nov
Dec
Jan
Feb
M ar
Apr
M ay
Jun
Jul
Aug
S ep
Oct
Nov
Dec
Jan
Feb
0
2004
2005
2006
2007
2008
Collection Period
Figure 2—Detection of P. ramorum by multiplex PCR from leaf baits in 32 streams
monitored continuously from March 2004 through February 2008. There were no
samples collected in September 2004.
371
GENERAL TECHNICAL REPORT PSW-GTR-229
Periods of low recovery frequency corresponded with cooler water temperatures (fig. 3) and
high winter flow rates in the streams (data not shown).
16
12
17
PCR positive
15
Mean water temp
13
8
11
4
9
0
7
Jan Feb Mar Apr May Jun
Jul
Mean Water
Temperature (C)
PCR Positive (%)
20
Aug Sep Oct Nov Dec
2006
Figure 3—Monthly multiplex PCR detection of P. ramorum in leaf baits from 32
streams monitored in 2006 and water temperature of those streams.
Acknowledgments
Thanks to the U.S. Department of Agriculture, Forest Service, Pacific Northwest Region
(Region 6) Forest Health Monitoring Program, and the Pacific Southwest Research Station for
funding.
372
Proceedings of the Sudden Oak Death Fourth Science Symposium
Factors Affecting Onset of Sporulation of
Phytophthora ramorum on Rhododendron
‘Cunningham’s White’1
Paul W. Tooley2 and Marsha E. Browning2
Abstract
Phytophthora ramorum is the Oomycete pathogen that causes sudden oak death (SOD). SOD
continues to cause heavy losses to California and Oregon forests as well as to the U.S. nursery
industry (Dart and Chastagner 2007; Frankel 2008), in part due to quarantine regulations.
Epidemics are driven by sporangia, which are produced in large numbers on various P.
ramorum hosts (Parke and others 2002; Tooley and Browning 2009). While some sporulation
has been observed on leaves of coast live oak (Quercus agrifolia) (Vettraino and others 2008),
California bay laurel (Umbellularia californica) is thought to be the major producer of
sporangia that drive forest epidemics (Davidson and others 2008). Little is known of the
parameters associated with the onset of sporangia production by P. ramorum on its hosts
following infection and on leaves with established disease. We conducted experiments to
examine the relationship between lesion size, moisture period, and temperature with the first
appearance of dehisced sporangia. Leaves exhibiting small lesions induced via artificial
inoculation with isolate CAM 5C using a variety of techniques were positioned on 15 µmmesh nylon screens in a mist chamber inside a greenhouse cubicle (15 to 18 °C). Lesions
were measured daily, and sporangia were collected and counted. Lesion size did not appear to
be a reliable predictor for onset of sporangia production. We observed a wide range of lesion
sizes (1 to 54 mm2) corresponding to the initial collection of dehisced sporangia from
diseased leaves. In contrast, the longer the moist period duration, the higher the percentage of
diseased leaves supporting sporangia production by P. ramorum – 17 to 22 percent within 24
hours, and an additional 39 to 64 percent within 48 hours. On leaves with established disease,
dehisced sporangia were collected from ≥ 50 percent of leaves within a 6- to 12-hour moist
period. In temperature studies, sporangia were collected from misted detached leaves within 1
day at 15 °C; 2 days at 10 °C and 20 °C; and 3 days at 4 °C, 25 °C, and 30 °C. When infected
leaves were first pre-incubated at 20 °C for 72 hours, allowing lesions to expand uniformly,
sporangia were collected from some leaves following a 1 day incubation at all temperatures.
Knowledge of the timing and conditions that allow P. ramorum to sporulate following host
infection will contribute to our understanding of epidemic dynamics and improve our ability
to predict the likelihood that P. ramorum may spread within forest and nursery environments
(Moralejo and others 2006).
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
USDA-ARS Foreign Disease Weed Science Research Unit, 1301 Ditto Ave., Fort Detrick, MD 21702.
Corresponding author: paul.tooley@ars.usda.gov.
373
GENERAL TECHNICAL REPORT PSW-GTR-229
Literature Cited
Dart, N.L. and Chastagner, G.A. 2007. Estimated economic losses associated with the
destruction of plants due to Phytophthora ramorum quarantine efforts in Washington State.
Online, Plant Health Progress doi: 10.1094/PHP-2007-0508-02-RS.
Davidson, J.M.; Patterson, H.A. and Rizzo, D.M. 2008. Sources of inoculum for
Phytophthora ramorum in a redwood forest. Phytopathology. 98:.860–866.
Frankel, S. 2008. Sudden oak death and Phytophthora ramorum in the USA: a management
challenge. Australasian Plant Pathology. 37: 19–25.
Moralejo, E.; García Muñoz, J.A. and Descals, E. 2006. Insights into Phytophthora
ramorum sporulation: epidemiological and evolutionary implications. OEPP/EPPO Bulletin
36:383–368.
Parke, J.L.; Hansen, E.M. and Linderman, R.G. 2002. Sporulation potential of
Phytophthora ramorum on leaf disks from selected hosts. Sudden oak death science
symposium, Monterey, CA, 15-18 December 2002.
http://danr.ucop.edu/ihrmp/sodsymp/paper/paper26.html.
Tooley, P.W. and Browning, M. 2009. Susceptibility to Phytophthora ramorum and
inoculum production potential of some common eastern forest understory plant species. Plant
Disease. 93:249–256.
Vettraino, A.M.; Huberli, D. and Garbelotto, M. 2008. Phytophthora ramorum infection of
coast live oak leaves in Californian forests and its capacity to sporulate in vitro. Australasian
Plant Pathology. 37:72–73.
374
Proceedings of the Sudden Oak Death Fourth Science Symposium
Screening Trichoderma asperellum as a
Mycoparasite on Phytophthora ramorum1
Timothy L. Widmer2 and Gary J. Samuels3
Abstract
Despite efforts of eradication and sanitation, Phytophthora ramorum persists in the United
States and abroad. Fungicides have limited effectiveness, but there are concerns that they may
only inhibit pathogen growth and resistance may develop. Biological control is an active
control measure that can work continuously as long as the agent is alive and active. The goal
of this study was to examine whether Trichoderma asperellum isolates are mycoparasitic on
P. ramorum. Sixteen isolates of T. asperellum and other Trichoderma spp. that have
demonstrated antagonism towards other Phytophthora spp., or were suspected to have
mycoparasitic activity, were selected. The rate of mycoparasitism was determined by
overlaying a strip of Trichoderma-colonized agar on a V8 agar plate colonized by P. ramorum
(A2 mating type). Every 7 days for 4 weeks agar plugs were removed and transferred to V8
agar amended with benomyl (V8+B) or a wounded leaf disk of Rhododendron
‘Cunningham’s White.’ Control plugs of P. ramorum, without exposure to the Trichoderma
spp., always showed growth on V8+B and produced necrosis on the leaf disks. The different
Trichoderma spp. isolates demonstrated variable mycoparasitic activities. Some isolates
showed no inhibition of P. ramorum growth on V8+B or reduction in necrosis even from
plugs removed directly below the Trichoderma strip. Other isolates showed a reduction in
growth and necrosis over time, but did not completely eliminate the pathogen after 4 weeks.
Six isolates of T. asperellum were consistent among replicated trials in eliminating growth of
P. ramorum from the agar plugs and preventing leaf disk necrosis within 2 weeks exposure.
Further testing of four of these six T. asperellum isolates against two different P. ramorum
isolates (A1 and A2 mating types) resulted in the same high level of mycoparasitic activity.
We believe the results demonstrate that specific T. asperellum isolates have the potential to
remediate P. ramorum-infested soil. Tests are ongoing to determine whether P. ramorum soil
populations can be eliminated when treated with these isolates.
Introduction
Studies have shown that Phytophthora ramorum can survive in potting medium
around containers of infected plants in a nursery (Jeffers 2005, Tjosvold and others
2009). Aerated steam and chemical fumigants are known methods to eliminate
soilborne pathogens. Linderman and Davis (2008) found that P. ramorum
populations in potting media were killed by aerated steam heat treatments of 50 oC or
higher or treatment with metam sodium concentrations of 0.25 ml per l of medium.
However, using these techniques increases the chance of destroying beneficial
microorganisms and of working in a hazardous environment.
1
A version of this paper was presented at the Fourth Sudden Oak Death Science Symposium, June 1518, 2009, Santa Cruz, California.
2
USDA/ARS, Foreign Disease and Weed Science Research Unit, 1301 Ditto Avenue, Fort Detrick, MD
21702.
3
USDA/ARS, Systematic Botany and Mycology Laboratory, Beltsville, MD 20705.
Corresponding author: tim.widmer@ars.usda.gov.
375
GENERAL TECHNICAL REPORT PSW-GTR-229
Some of the most studied and promising fungi used in a biocontrol system are
Trichoderma spp. (Harman and others 2004). Populations of Trichoderma spp.,
which often are abundant in composts and compost-amended media, typically
suppress Pythium and Phytophthora root rots within days after their formulation (De
Ceuster and Hoitink 1999). Trichoderma spp. are reported to suppress soilborne
diseases caused by Phytophthora spp. in containerized systems (da S. Costa and
others 2000, Sharifi Tehrani and Nazari 2004). Currently, T. asperellum is being
studied as a biological control agent to manage black pod disease of cacao in
Cameroon. Recent results show that disease incidence was lower when T. asperellum
was applied on infected cacao trees (Tondje and others 2007). It was the purpose of
this study to screen selected Trichoderma spp. for antagonism towards P. ramorum.
Materials and Methods
Sixteen different Trichoderma spp. isolates were cultured on half-strength PDA
(1/2PDA). This included 12 isolates of T. asperellum, two isolates of T. virens, one
isolate of T. koningiopsis, and one undescribed isolate Trichoderma sp. nov. Three
different P. ramorum isolates, WSDA-1772, 5-C, both A2 mating type and clonal
lineage NA1, and PRN-1 (CBS 101327), mating type A1 and clonal lineage EU1,
were cultured on 20 percent clarified V8 agar.
An agar plate bioassay was conducted as described by Krauss and others (1998). The
Trichoderma spp. were grown on 1/2PDA in 90 mm diameter Petri plates until they
completely colonized the plate. A 4 X 1 cm strip of Trichoderma-colonized 1/2PDA
was removed and transferred to a P. ramorum-colonized V8 agar plate. A noncolonized 1/2PDA strip was used in the same way as a control. Every week for 4
weeks a 1 cm X 4.5 cm strip perpendicular to the original Trichoderma strip was
removed. The strip was cut lengthwise in half and divided into 0.5 cm cubes to give
two sets of nine cubes. From one of the sets the cubes were placed individually on the
abaxial side of nine wounded Rhododendron ‘Cunningham’s White’ leaf disks (6 mm
diameter). The corresponding set was placed on a 20 percent V8 agar plate
supplemented with 50 mg/l of benomyl. After 1 week at 20oC, observations were
made on the leaf disks for necrosis and mycelial growth originating from the cubes
on the V8 agar plate. The experiment was conducted twice for each Trichoderma
isolate on the three different P. ramorum isolates.
Results
There was an observed correlation between the lack of necrosis on the leaf disks and
no mycelial growth from the corresponding plug on the agar plate. Nine Trichoderma
spp. isolates showed some reduction in necrosis after 1 week and complete reduction
in necrosis and P. ramorum growth within 2 weeks after exposure. Eight of these
isolates were T. asperellum and the other one was T. koningiopsis. There was no
difference among the P. ramorum isolates tested.
Microscopic examination of the interaction between the antagonistic Trichoderma
spp. and P. ramorum revealed mycoparasitism of P. ramorum chlamydospores and
sporangia (fig. 1). This confirms the observation by Watanabe and others (2007),
who first reported mycoparasitism as the mode of action of T. asperellum.
376
Proceedings of the Sudden Oak Death Fourth Science Symposium
A
B
Figure 1—Trichoderma asperellum (arrows) demonstrating mycoparasitism of a P.
ramorum A) chlamydospore and B) sporangium. Bar = 20 μm.
Discussion
Specific T. asperellum and T. koningiopsis isolates have potential as biocontrol
agents against P. ramorum. Further tests have been started to determine if selected
isolates can reduce P. ramorum soil populations to nondetectable levels over time. In
addition, tests are ongoing to determine if selected isolates can parasitize and
eliminate viability of P. ramorum propagules in infested leaf litter. If the
Trichoderma spp. are found to be effective in eliminating the population of P.
ramorum in the soil and infested leaf litter, then the potential exists to use this as a
biologically based method to remediate infested nursery beds.
377
GENERAL TECHNICAL REPORT PSW-GTR-229
Literature Cited
da S. Costa, J.L.; Menge, J.A. and Casale, W.L. 2000. Biological control of Phytophthora
root rot of avocado with microorganisms grown in organic mulches. Brazilian Journal of
Microbiology. 31: 239–246.
De Ceuster, T.J.J. and Hoitink, H.A.J. 1999. Prospects for composts and biocontrol agents
as substitutes for methyl bromide in biological control of plant diseases. Compost Science and
Utilization. 7: 6–15.
Jeffers, S. 2005. Recovery of Phytophthora ramorum from soilless mixes around containergrown ornamental plants. Phytopathology. 95: S48.
Krauss, U.; Bidwell, R. and Ince, J. 1998. Isolation and preliminary evaluation of
mycoparasites as biocontrol agents of crown rot of banana. Biological Control. 13: 111–119.
Linderman, R.G. and Davis, E.A. 2008. Eradication of Phytophthora ramorum and other
pathogens from potting medium or soil by treatment with aerated steam or fumigation with
metam sodium. HortTechnology. 18: 106–110.
Sharifi Tehrani, A. and Nazari, S. 2004. Antagonistic effects of Trichoderma harzianum on
Phytophthora drechsleri, the casual agent of cucumber damping-off. Acta Horticulturae. 635:
137–139.
Tjosvold, S.A.; Chambers, D.L.; Fichtner, E.J. [and others]. 2009. Disease risk of potting
media infested with Phytophthora ramorum under nursery conditions. Plant Disease. 93: 371–
376.
Tondje, P.R.; Roberts, D.P.; Bon, M.C. [and others]. 2007. Isolation and identification of
mycoparasitic isolates of Trichoderma asperellum with potential for suppression of black pod
disease of cacao in Cameroon. Biological Control. 43: 202–212.
Watanabe, S.; Kumakura, K.; Izawa, N. [and others]. 2007. Mode of action of
Trichoderma asperellum SKT-1, a biocontrol agent against Gibberella fujikuroi. Journal of
Pesticide Science. 32: 222–228.
378